Nemotron-Parse#

Nemotron-Parse-v1.1 is NVIDIA’s document parsing VLM, specializing in extracting structured information from complex documents including tables, forms, and mixed-content PDFs.

Task

Document Parsing

Architecture

NemotronParseForConditionalGeneration

Parameters

varies

HF Org

nvidia

Available Models#

  • Nemotron-Parse-v1.1

Architecture#

  • NemotronParseForConditionalGeneration

Example HF Models#

Model

HF ID

Nemotron-Parse v1.1

nvidia/Nemotron-Parse-v1.1

Example Recipes#

Recipe

Dataset

Description

Try on Brev

nemotron_parse_v1_1.yaml

cord-v2

SFT — Nemotron-Parse on CORD-v2

Launch on Brev

Try with NeMo AutoModel#

1. Install (full instructions):

pip install nemo-automodel

2. Clone the repo to get the example recipes:

git clone https://github.com/NVIDIA-NeMo/Automodel.git
cd Automodel

3. Run the recipe from inside the repo:

automodel --nproc-per-node=8 examples/vlm_finetune/nemotron/nemotron_parse_v1_1.yaml
Run with Docker

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:26.02.00

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

cd /opt/Automodel

3. Run the recipe:

automodel --nproc-per-node=8 examples/vlm_finetune/nemotron/nemotron_parse_v1_1.yaml

See the Installation Guide and VLM Fine-Tuning Guide.

Fine-Tuning Tutorial on Brev#

Launch the end-to-end Nemotron Parse fine-tuning tutorial on Brev with a single click:

Launch on Brev

See also the tutorial notebook and the VLM Fine-Tuning Guide.

Hugging Face Model Cards#