Nodes/ComfyUI-WanVideoWrapper_QQ/Lora Extract K Jv2 (QQ)
ComfyUI Node

Lora Extract K Jv2 (QQ)

A ComfyUI node in KJNodes/lora with 17 inputs and 0 outputs.

By siraxe·Created 10 months ago·Updated 4 months ago· 71
Lora Extract K Jv2 (QQ)
  • finetuned_model
  • original_model
    filename_prefixloras/ComfyUI_v2_extracted_lora
    rank8
    lora_type
    algorithmsvd_linalg
    lowrank_iters7
    output_dtypefp16
    bias_difftrue
    adaptive_param0.15
    clamp_quantiletrue
    vram_usage_modebalanced
    blocks_in_gpu10
    batch_size4
    prefetch_ahead2
    memory_threshold0.85
    debug_modefalse
    CategoryKJNodes/lora

    Inputs (17)

    NameTypeDefaultDescription
    finetuned_modelMODEL
    original_modelMODEL
    filename_prefixSTRINGloras/ComfyUI_v2_extracted_lora
    rankINT81–4096The rank to use for standard LoRA, or maximum rank limit for adaptive methods.
    lora_typeCOMBO6 options: standard, full, adaptive_ratio, adaptive_quantile, adaptive_energy, adaptive_fro
    algorithmCOMBOsvd_linalgSVD algorithm to use, svd_lowrank is faster but less accurate.
    lowrank_itersINT71–100The number of subspace iterations for lowrank SVD algorithm.
    output_dtypeCOMBOfp163 options: fp16, bf16, fp32
    bias_diffBOOLEANtrue
    adaptive_paramFLOAT0.150–1For ratio mode, this is the ratio of the maximum singular value. For quantile mode, this is the quantile of the singular values. For fro mode, this is the Frobenius norm retention ratio.
    clamp_quantileBOOLEANtrue
    vram_usage_modeCOMBObalancedVRAM usage mode. Conservative=50% VRAM, Balanced=75% VRAM, Aggressive=90%+ VRAM
    blocks_in_gpuINT100–40Number of transformer blocks to keep in GPU memory (0=offload all, 40=keep all)
    batch_sizeINT41–8Number of layers to process simultaneously (higher=more VRAM, faster)
    prefetch_aheadINT20–10Number of blocks to prefetch ahead (reduces I/O bottlenecks)
    memory_thresholdFLOAT0.850.5–0.95VRAM usage threshold before automatic offloading (0.5-0.95)
    debug_modeBOOLEANfalseEnable detailed performance and VRAM usage debugging

    Outputs (0)

    No outputs