Model CoverageVision Language Models

Mistral-Small-4

View as Markdown

Mistral-Small-4-119B is Mistral AI’s multimodal MoE model supporting both text and image inputs at scale.

TaskImage-Text-to-Text
ArchitectureMistralForConditionalGeneration
Parameters119B (MoE)
HF Orgmistralai

Available Models

  • Mistral-Small-4-119B-2603

Architecture

  • MistralForConditionalGeneration

Example HF Models

ModelHF ID
Mistral-Small-4 119Bmistralai/Mistral-Small-4-119B-2603

Example Recipes

RecipeDatasetDescription
mistral4_medpix.yamlMedPix-VQASFT — Mistral-Small-4 on MedPix

Try with NeMo AutoModel

1. Clone and install from source (full instructions):

$git clone https://github.com/NVIDIA-NeMo/Automodel.git
$cd Automodel
$uv sync --locked --all-groups --all-extras --extra vlm-media

This recipe was validated on 4 nodes × 8 GPUs (32 H100s). See the Launcher Guide for multi-node setup.

2. Run the recipe from inside the repo:

$uv run automodel --nproc-per-node=8 examples/vlm_finetune/mistral4/mistral4_medpix.yaml

1. Pull the container and mount a checkpoint directory:

$docker run --gpus all -it --rm \
>  --shm-size=8g \
>  -v $(pwd)/checkpoints:/opt/Automodel/checkpoints \
>  nvcr.io/nvidia/nemo-automodel:26.06.00

2. Navigate to the AutoModel directory (where the recipes are):

$cd /opt/Automodel

3. Install media dependencies (not installed in the container by default):

$uv pip install ".[vlm-media]"

4. Run the recipe:

$automodel --nproc-per-node=8 examples/vlm_finetune/mistral4/mistral4_medpix.yaml

See the Installation Guide and VLM Fine-Tuning Guide.

Fine-Tuning

See the VLM Fine-Tuning Guide.

Hugging Face Model Cards