TSM2330 Style - MonsieurBaka

☕ Support My Work: If you enjoy my local tools and models and want to support me with real cents rather than just platform Buzz, you can do so here: https://ko-fi.com/monsieurbaka. Your support directly speeds up new software updates and model releases!

Core Trigger Tags & Prompting

To trigger the aesthetic, use:

• score_9, masterpiece, @tsm2330 style or simply @tsm2330 style

Automation & Workflows (The PromptForge Ecosystem)

To save you from manually copying and pasting style tags or tweaking configuration JSONs, I am releasing the official PromptForge library integration alongside this model:

• [HUGGINGFACE LINK] Official PromptForge Library Pack addon for this style (.txt, .jpg, .json)

For those who value their time and want to streamline multi-character prompting alongside complex style LoRA management directly in ComfyUI:

PromptForge is a standalone software designed to automate prompting workflows and seamlessly handle complex LoRA dependencies during image generation. The brand-new major release is faster, more functional, and completely overhauled for max workflow efficiency. Read the full changelog, guide, and download instructions in the official article:

👉 https://civitai.red/articles/32428/promptforge-2-automatization-strikes-back

tsm2330 Art Style

Showcase Setup & Model Weight

• Recommended LoRA Weight: 1.0

• Showcase Setup: The entire showcase gallery on this page was generated at a solid weight of 1.0 for this Style LoRA (and 1.0 for character LoRAs whenever paired).

We aren't hiding behind low weights here - this page is a raw flex to demonstrate that the model has deeply baked the artist's unique lighting dynamics, volumetric depth, and material rendering directly into its weights without collapsing the underlying geometry.

About the Artist & Style Breakdown

This model is a dedicated style LoRA trained on the public artwork dataset of the talented neural artist tsm2330. In my opinion, tsm2330 has achieved the ultimate sweet spot in anime-style rendering: a perfect balance between sharp, aggressive rim lighting, soft shadows, and vibrant subsurface scattering (SSS) for skin materials.

Dataset Architecture & Captioning (Shoutout to Grok)

• Dataset Size: 250 surgically clean images, all manually curated, hand-filtered, and carefully pre-processed.

• The Captioning MVP: Credit where credit is due - a huge chunk of the detailed captions was generated with massive help from Grok. When it comes to unhinged, uncensored, and detailed description of spicy/NSFW visual content, Grok remains the only capable non-human assistant on the market willing to write captions without throwing a censorship fit.

• NSFW Bias: Due to the specific nature of tsm2330’s portfolio, a significant portion of the dataset consists of NSFW material. As a result, the model possesses a strong natural bias toward realistic skin rendering and detailed anatomy. Keep this in mind when crafting your prompts!

• Zero Concept Bleeding: Despite the high stylistic dominance, all recurring characters inside the training set were isolated on a semantic token level. Calling a specific lighting setup or pose tag will never force an unwanted character's features or hair traits into your generation.