CosmicLaca-ComfyUI_Primere_Nodes/README.md

Primere nodes for ComfyUI

Git link: https://github.com/CosmicLaca/ComfyUI_Primere_Nodes

---

Nodepack under development. Manual will be ready later. If you ugrade just check attached workflows or use git to downgrade to previous version.

---

All workflows in the Workflow folder of the node root.

---

Minimal workflow:

Included features:

• Prompt selector to any prompt sources
• CSV and TOML file source readers for saved prompts, automatically organized, saved prompt selection by preview image
• Randomized latent noise for variations
• Prompt encoder with selectable custom clip model, long-clip mode with custom models, advanced encoding, injectable internal styles, last-layer options
• Sampler with variation extender and Align Your Step
• A1111 style network injection supported by text prompt (Lora, Lycorys, Hypernetwork, Embedding)
• Automatized and manual image saver. Manual image saver with optional preview saver for checkpoint selector and saved CSV prompts
• Upscaler (Ultimate SD and hiresFix)

Examples:

Visual checkpopint selection, automatized filtering by subdirectories (first row of buttons) and versions (second row of buttons):

Visual saved prompt selection (csv source), automatized filtering by categories:

---

Basic workflow:

Included features:

Same as Minimal workflow plus:

• Half-automatic concept selector:

  • Supported concepts: SD1, SD2, SDXL, SD3, StableCascade, Turbo, Flux, KwaiKolors, Hunyuan, Playground, Pony, LCM, Lightning, Hyper
  • Custom sampler settings for all supported concepts
  • Auto detection of selected model type (if data already stored)
  • Auto download and apply HyperSD speed loras at first usage from here: https://huggingface.co/ByteDance/Hyper-SD/tree/main check your SSD space before!

• Terminal helper to detect and store model concept data:

  • Open cmd terminal window
  • Activate your Comfy venv
  • Change to [Your_comfy_folder]\custom_nodes\ComfyUI_Primere_Nodes\terminal_helpers\
  • Enter model_version_cache.py

• You will see like this when run:

• Terminal helper cannot detect all your models correctly, specially your symlinks, because diffusers. You must check this file: Nodes/.cache/.cache.json and seek the word symlink and unknown
• Manually replace these bad results on .cache.json to right value. If you don't use symlinks, the version info will much more correct.
• Refresh your browser, and you will see version info on the top left badge of preview (instead of questionmark)

When all data available, thse badges will visible in the preview:

• Top left: model concept (Flux, SDXL, etc...)
• Top right: if symlinked, what type of diffuser
• Botom: the average aesthetic score. Have to use aesthetic scorer node before to store this data for checkpoints or saved prompts. The number is the average, but the percent depending on the checkpoint selector settings, where the aescore_percent_min and lower value mean 0%, aescore_percent_max and higher mean 100%.

On the Concept selector node you will see None on all required fields, for example on Cascade files. You must install all used concept files manually to right path, then select correct models on these lists.

When you open the workflow first time:

When downloaded required files set to: