Devstral-2-123B-Instruct-2512 | NVIDIA NeMo AutoModel

Devstral-2-123B-Instruct-2512 has a checked-in NeMo AutoModel recipe for text generation. The Hugging Face configuration declares the Ministral3ForCausalLM architecture.

Fine-Tune Devstral-2-123B-Instruct-2512

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

automodel examples/llm_finetune/devstral/devstral2_123b_2512_squad_peft.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=8, pp_size=1, cp_size=1, ep_size=1
Nodes	1
Features	Checkpointing (enabled=true), Activation checkpointing (activation_checkpointing=true), LoRA
Advanced	bfloat16

Model Reference

Model Architecture

Property	Value
Task	Text generation
Architecture	`Ministral3ForCausalLM`

Available Models

mistralai/Devstral-2-123B-Instruct-2512

LLM fine-tuning guide