TencentQQGYLab-ComfyUI-ELLA/NODES.md

Nodes reference

Loaders

Load ELLA Model

Configuration parameters

• name: (STRING), ELLA model to load.

Outputs

• ella: (ELLA), ELLA model.

Load T5 TextEncoder #ELLA

Configuration parameters

• name: (STRING), T5 text encoder model to load.
• max_length: (INT), max_length of encoder, set 0 for flexible token length.
• dtype: (BOOLEAN)

Outputs

• t5_encoder: (T5_TEXT_ENCODER)

Main Apply Nodes

Apply ELLA

Inputs

• model: (MODEL), loaded by Load Checkpoint and other model loaders.
• ella: (ELLA), ELLA model loaded by Load ELLA Model node.
• positive: (ELLA_EMBEDS), currently needing both.
• negative: (ELLA_EMBEDS), currently needing both.

Optional

• sigmas: (SIGMAS), Use BasicScheduler and same config with KSampler to get sigmas. Fallback into legacy method(deprecated) if not provided and timesteps is not set.

Outputs

• model: (MODEL), sd model with ELLA model injected.
• positive: (CONDITIONING), for KSamplers.
• negative: (CONDITIONING), for KSamplers.

ELLA Encode

Inputs

• ella: (ELLA), ELLA model loaded by Load ELLA Model node. Need to use Set ELLA Timesteps first.
• embeds: (ELLA_EMBEDS)

Outputs

• conds: (CONDITIONING), for KSamplers.

ELLA Text Encode

If clip is provided to this node, it will automatically concat ella condition and clip condition.

Inputs

• ella: (ELLA), ELLA model loaded by Load ELLA Model node. Need to use Set ELLA Timesteps first.
• text_encoder: (T5_TEXT_ENCODER)
• text: (STRING), prompt to encode.

Optional

• clip: (CLIP), clip model to encode text_clip.
• text_clip: (STRING), prompt to encode with clip.

Outputs

• conds: (CONDITIONING), for KSamplers.

T5 Text Encode #ELLA

Inputs

• text: (STRING), prompt to encode.
• text_encoder: (T5_TEXT_ENCODER)

Optional

• embeds: (ELLA_EMBEDS), for chain combination ELLA_EMBEDS

Outputs

• ella_embeds: (ELLA_EMBEDS)

Helpers

ELLA Combine Embeds

Inputs

• embeds: (ELLA_EMBEDS)
• embeds_add: (ELLA_EMBEDS)

Outputs

• embeds: (ELLA_EMBEDS)

Convert Condition to ELLA Embeds

Deprecated, 'Combine CLIP & ELLA Embeds' instead

use for convert clip CONDITIONING to ELLA_EMBEDS

Inputs

• cond: (CONDITIONING)

Outputs

• embeds: (ELLA_EMBEDS)

Concat Condition & ELLA Embeds

Deprecated, 'Combine CLIP & ELLA Embeds' instead

Combine CLIP & ELLA Embeds

Inputs

• cond: (CONDITIONING)
• embeds: (ELLA_EMBEDS)

Outputs

• embeds: (ELLA_EMBEDS)

Set ELLA Timesteps

Inputs

• model: (MODEL), loaded by Load Checkpoint and other model loaders.
• ella: (ELLA), ELLA model loaded by Load ELLA Model node.
• scheduler: (SCHEDULER_NAMES), need to be consistent with Ksamper.
• steps: ("INT"), need to be consistent with Ksamper.
• denoise: ("FLOAT"), need to be consistent with Ksamper.

Outputs

• ella: (ELLA), ELLA model with timesteps injected.