nemotron_h

Model type	nemotron_h
Class	NemotronHCausalLMModel
Task	hybrid-text-generation
Source	models/nemotron_h.py

Description

NemotronH hybrid Mamba2+Attention+MLP causal language model.

Uses HybridCausalLMTask with mixed "mamba2", "full_attention", and "mlp" layer types for the cache.

HuggingFace reference: NemotronHForCausalLM.

Usage

mobius build --model <MODEL_ID> output_dir/

from mobius import build

model = build("<MODEL_ID>")