Nodes/DiffusionGemma Prompt Builder/DiffusionGemma Model Loader (Advanced)
ComfyUI Node

DiffusionGemma Model Loader (Advanced)

A ComfyUI node in prompt/diffusiongemma with 8 inputs and 2 outputs.

By exportAnything·Created about a month ago·Updated about a month ago· 43
DiffusionGemma Model Loader (Advanced)
    • model_config
    • runtime_status_json
    model_path/tmp/ComfyUI/custom_nodes/models/LLM/diffusiongemma-26B-A4B-it
    backendtransformers_inprocess
    dtypeauto
    quantizationmodelopt_nvfp4
    local_files_onlytrue
    unload_policykeep_loaded
    max_memory_gb20.0 GB - safe practical
    temperature0.45
    Categoryprompt/diffusiongemma

    Inputs (8)

    NameTypeDefaultDescription
    model_pathSTRING/tmp/ComfyUI/custom_nodes/models/LLM/diffusiongemma-26B-A4B-itWhole Hugging Face repo folder for the in-process Transformers/NVFP4 runtime.
    backendCOMBOtransformers_inprocess1 options: transformers_inprocess
    dtypeCOMBOauto4 options: auto, bfloat16, float16, float32
    quantizationCOMBOmodelopt_nvfp45 options: modelopt_nvfp4, none, bitsandbytes_4bit, quanto, nvfp4_metadata_only
    local_files_onlyBOOLEANtrue
    unload_policyCOMBOkeep_loaded2 options: keep_loaded, unload_after_run
    max_memory_gbCOMBO20.0 GB - safe practicalNVFP4 bridge memory profile. 20 GB is the safest practical default; 18 GB is the supported floor; 0 requests full GPU load.
    temperatureoptFLOAT0.450–2Generation sampling temperature. 0.2 is conservative; 0.45-0.8 is usually more creative.

    Outputs (2)

    NameTypeDescription
    model_configDG_MODEL_CONFIG
    runtime_status_jsonSTRING