Skip to main content
Ctrl+K
NeMo-AutoModel - Home NeMo-AutoModel - Home

NeMo-AutoModel

  • GitHub
NeMo-AutoModel - Home NeMo-AutoModel - Home

NeMo-AutoModel
  • GitHub

Table of Contents

Get Started

  • About
  • Key Features
  • Installation
  • Configuration
  • 🤗 HF Compatibility
  • Repo Structure
  • Release Notes

Announcements

  • Announcements

Performance

  • Performance Summary

Model Coverage

  • Overview
  • Release Log
  • Large Language Models
    • Llama
    • Gemma
    • Qwen2
    • Qwen2 MoE
    • Qwen3
    • Qwen3 MoE
    • Qwen3-Next
    • ERNIE 4.5
    • DeepSeek
    • DeepSeek-V3
    • DeepSeek V4 Flash
    • Mistral
    • Mixtral
    • Ministral3 / Devstral
    • Phi
    • Phi-3 / Phi-4
    • Phi-3-Small
    • Nemotron / Minitron
    • Nemotron-H
    • Nemotron-Flash
    • Nemotron-Super (Llama-3.3-Nemotron-Super-49B)
    • ChatGLM
    • GLM-4
    • GLM-4 MoE (GLM-4.5 / GLM-4.7)
    • GLM-5 / GLM-5.1 (MoE + DSA)
    • Granite
    • Granite MoE
    • Bamba
    • OLMo
    • OLMo2
    • OLMoE
    • GPT-OSS
    • GPT-2
    • GPT-J
    • GPT-NeoX / Pythia
    • StarCoder
    • StarCoder2
    • Aquila / Aquila2
    • Baichuan / Baichuan2
    • Command-R
    • Falcon
    • EXAONE
    • InternLM
    • Jais
    • MiniMax-M2
    • MiniCPM
    • Moonlight
    • Seed (ByteDance)
    • Solar Pro
    • Orion
    • StableLM
    • Step-3.5
    • GritLM
    • Hy3 (HunyuanLarge)
    • Hy-MT2 (Hunyuan-MT2)
    • MiMo-V2-Flash
    • Ling 2.0
  • Vision Language Models
    • Kimi-VL
    • Gemma 3 VL / Gemma 3n
    • Gemma 4
    • Qwen2.5-VL
    • Qwen3-VL / Qwen3-VL-MoE
    • Qwen3.5-VL
    • Nemotron-Parse
    • Ministral3 VL
    • Mistral-Small-4
    • Mistral Medium 3.5
    • InternVL
    • Llama 4
    • SmolVLM
    • LLaVA
    • LLaVA-OneVision 1.5
    • Step-3.7-Flash
  • Omni
    • Qwen3-Omni
    • Qwen2.5-Omni
    • Phi-4-multimodal
    • Nemotron-3-Nano-Omni
  • Diffusion
    • Wan 2.1 T2V
    • FLUX.1-dev
    • HunyuanVideo 1.5
    • Qwen-Image
  • Embedding Models
    • Llama (Bidirectional) for Embedding
    • Ministral3 (Bidirectional) for Embedding
  • Reranking Models
    • Llama (Bidirectional) for Reranking

Recipes & E2E Examples

  • Recipes and End-to-End Examples
  • SFT & PEFT
  • Function Calling
  • Knowledge Distillation
  • Large MoE Fine-Tuning
  • DeepSeek V4 Flash
  • Hy3-preview
  • Pretraining
  • NanoGPT Pretraining
  • Sequence Classification
  • Gemma 3 / 3n
  • Gemma 4
  • Qwen3.5-VL
  • Step-3.7-Flash
  • Nemotron-Omni
  • Mistral Medium 3.5 VL
  • Qwen3-Omni ASR
  • Diffusion Fine-Tuning
  • dLLM Fine-Tuning
  • QAT
  • Databricks
  • SGLang EAGLE Serving

Datasets

  • Overview
  • Text Dataset
  • Retrieval Dataset
  • ColumnMapped Dataset
  • ColumnMapped Iterable
  • Multi-Modal Dataset
  • Diffusion Dataset

Job Launchers

  • Overview
  • Local Workstation
  • SLURM Cluster
  • NeMo Run
  • SkyPilot
  • SkyPilot k8s

Development

  • Checkpointing
  • Gradient Checkpointing
  • Pipeline Parallelism
  • Mixed-Precision Training
  • FP8 Training
  • MLflow Logging
  • API Reference
    • nemo_automodel
      • nemo_automodel._diffusers
        • nemo_automodel._diffusers.auto_diffusion_pipeline
      • nemo_automodel._transformers
        • nemo_automodel._transformers.v4_patches
        • nemo_automodel._transformers.model_init
        • nemo_automodel._transformers.mfu
        • nemo_automodel._transformers.utils
        • nemo_automodel._transformers.infrastructure
        • nemo_automodel._transformers.kernel_patches
        • nemo_automodel._transformers.auto_tokenizer
        • nemo_automodel._transformers.registry
        • nemo_automodel._transformers.capabilities
        • nemo_automodel._transformers.auto_model
        • nemo_automodel._transformers.retrieval
        • nemo_automodel._transformers.te_attention
      • nemo_automodel.cli
        • nemo_automodel.cli.app
        • nemo_automodel.cli.utils
      • nemo_automodel.autonvtx
      • nemo_automodel.shared
        • nemo_automodel.shared.transformers_patches
        • nemo_automodel.shared.utils
        • nemo_automodel.shared.import_utils
        • nemo_automodel.shared.te_patches
        • nemo_automodel.shared.torch_patches
      • nemo_automodel.components
        • nemo_automodel.components.utils
        • nemo_automodel.components.attention
        • nemo_automodel.components.loss
        • nemo_automodel.components.flow_matching
        • nemo_automodel.components.training
        • nemo_automodel.components.config
        • nemo_automodel.components.moe
        • nemo_automodel.components.checkpoint
        • nemo_automodel.components.speculative
        • nemo_automodel.components.datasets
        • nemo_automodel.components.distributed
        • nemo_automodel.components._peft
        • nemo_automodel.components.quantization
        • nemo_automodel.components.models
        • nemo_automodel.components.optim
        • nemo_automodel.components.launcher
        • nemo_automodel.components.loggers
  • Breaking Changes
  • API Reference
  • nemo_automodel
  • nemo_automodel.components
  • nemo_automodel.components.models
  • nemo_automodel.components.models.hy_mt2

nemo_automodel.components.models.hy_mt2#

Submodules#

  • nemo_automodel.components.models.hy_mt2.layers
  • nemo_automodel.components.models.hy_mt2.dispatch
  • nemo_automodel.components.models.hy_mt2.model
  • nemo_automodel.components.models.hy_mt2.config
  • nemo_automodel.components.models.hy_mt2.state_dict_adapter

Package Contents#

Data#

__all__

API#

nemo_automodel.components.models.hy_mt2.__all__#

[‘HyMT2ForCausalLM’]

previous

nemo_automodel.components.models.llama.state_dict_adapter

next

nemo_automodel.components.models.hy_mt2.layers
On this page
  • Submodules
  • Package Contents
    • Data
    • API
      • __all__
NVIDIA NVIDIA
Privacy Policy | Manage My Privacy | Do Not Sell or Share My Data | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2026, NVIDIA Corporation.

Last updated on 2026-05-29.