Model CoverageVision Language Models

MuseGlimmer

View as Markdown

Muse-Glimmer-30B is a dense vision-language model with a 52-layer language backbone, a 50-layer vision tower, and a multimodal projector.

TaskImage-Text-to-Text
ArchitectureMuseGlimmerForConditionalGeneration
Model typesmuse_glimmer, muse_glimmer_text, muse_glimmer_vision
Parameters30B
HF IDmeta-models/Muse-Glimmer-30B

Architecture

  • Dense grouped-query attention language model with 32 query heads and 2 key-value heads
  • Alternating sliding and full attention across a 52-layer text backbone
  • Vision transformer with window and full attention across 50 layers
  • One-dimensional position IDs for packed Transformer Engine context parallelism

Example Recipes

RecipeDatasetDescription
muse_glimmer_30b_medpix.yamlMedPix-VQAFull-parameter VLM SFT
muse_glimmer_30b_medpix_lora.yamlMedPix-VQASingle-node LoRA SFT
muse_glimmer_30b_tulu3_te_tp2_cp4_packed_16k.yamlTulu 3Single-node 16K packed text SFT with TP2 and CP4

Try with NeMo AutoModel

The recipes use the Hugging Face checkpoint meta-models/Muse-Glimmer-30B by default:

$uv run automodel --nproc-per-node=8 \
>  examples/vlm_finetune/muse_glimmer/muse_glimmer_30b_medpix.yaml

The implementation supports FSDP2, activation checkpointing, TP1/TP2, context parallelism, Transformer Engine packed THD inputs, pipeline parallelism, and LoRA. Multi-axis mRoPE with packed THD context parallelism is not supported.