--Qwen2.1 Me recommendations: 3-4 cfg (but if image become slop, then change to 1) beta scheduler, res_2m / res_2s / res_multistep (2s best quality, multistep the fastest, 2m balance) I'll post my flow a bit later when finish uploading other loras P.S: im in a hurry so more exmaples will be soon for all my loras --Krea2 Actually i tested with different combinations raw and turbo. For best quality i use raw with cfg4, for fast result i use raw, but witth cfg1 and turbo lora (it really gives better quality then kreo2 turbo model). U can use with other loras freely, also works good with nicegirls lora and nsfw loras. recommended weight 0.8 for raw and 1.2-2 for turbo --Ideogram4 Works good with <a target="_blank" rel="ugc" href="https://civarchive.com/models/2688234/realism-engine-ideogram-4?modelVersionId=3026632">realism engine lora</a> for nsfw images. Also updated my repo for better lora control <a target="_blank" rel="ugc" href="https://github.com/DanrisiUA/ComfyUI-LoRA-Block-Filter">https://github.com/DanrisiUA/ComfyUI-LoRA-Block-Filter</a> One more. Made cool nodes with cursor (i have no clue how some of them to use) for ideogram <a target="_blank" rel="ugc" href="https://github.com/DanrisiUA/ComfyUI-Ideogram4-Enhanced.git">https://github.com/DanrisiUA/ComfyUI-Ideogram4-Enhanced.git</a>--Anima1.0Trained on top of my last <code>ultrareal_finetune_base1</code> checkpoint for a slight realism boost. I also "vibecoded" a few experimental nodes for Anima (pretty sure something similar already exists, but here we are): <a target="_blank" rel="ugc" href="https://github.com/DanrisiUA/ComfyUI-AnimaDynamicCFG">https://github.com/DanrisiUA/ComfyUI-AnimaDynamicCFG</a>Consider this a temporary realism upgrade until I train the next major fine-tune. --Anima0.5 Trained on anima3. Recommended strength: 0.6.Works best with block filtering — I created a ComfyUI nodepack for this: <a target="_blank" rel="ugc" href="https://github.com/DanrisiUA/ComfyUI-LoRA-Block-Filter">ComfyUI-LoRA-Block-Filter</a>. Examples generated alongside my other LoRA — nicegirls. Experimental version. Results may vary depending on prompt and block weight settings. --Flux2 I use it with res2m+beta57, guidance 4, and it works very well. The quality is similar to my Qwen Lenovo, but I noticed that Flux2 is more flexible and has better prompt adherence. Recommend to use with strength 0.85-0.9 -- Z-image Turbo Actually i'm not sure that my workflow is correct, i believe the result could be better Recommend to use with strength 0.80-0.9 I used euler + sgm_uniform with 4 guidance P.S: at this moment for ComfyUI u need this pack to load lora properly <a target="_blank" rel="ugc" href="https://github.com/PGCRT/CRT-Nodes?tab=readme-ov-file">https://github.com/PGCRT/CRT-Nodes</a> --Chroma My first try to train lora that enhance realism in Chroma. I'll upload my workflow today a lil bit later (or atm u can just take it from images) cause there some nuances about settings. But from my observations i noticed that for fast gen dpmpp2m + beta with cfg 3 works good . More experimental settigns are res_2s_rkmk2e + beta57 and cfg 3, more accurate gens, better limbs and fingers, but gives a bit weird grain on very small details + plus this method is pretty long --Qwen Works perfectly and much better then other versions. Gives extremely "live" images with amateurish quality with dramatic light and shadows, some blur and with vibe of 2000s street photography You can take workflow here <a target="_blank" rel="ugc" href="https://huggingface.co/Danrisi/Lenovo_Qwen/resolve/main/Qwen_danrisi.json">https://huggingface.co/Danrisi/Lenovo_Qwen/resolve/main/Qwen_danrisi.json</a> Thanx to <a target="_blank" rel="ugc" href="https://civarchive.com/user/fox23vang226">fox23vang226</a>. i understood that it's possible (and even have better effect) to use lora with values 1.5 - 2.0. P.S.: almost all example images generate with both loras (NiceGirls) --Wan This is the first release — mainly a test to explore Wan's capabilities. Note: In this initial version, facial details are somewhat lacking. For better results, I recommend using it alongside the<a target="_blank" rel="ugc" href="https://civarchive.com/models/1822984?modelVersionId=2066474"> Instagirls LoRA</a> by my friend, which significantly enhances facial features. --Flux Who needs a fancy name when the shadows and highlights do all the talking? This experimental LoRA is the scrappy cousin of my Samsung one—same punchy light-and-shadow mojo, but trained on a chaotic mix of pics from my ancient phones (so no Samsung for now)

Lenovo UltraReal

lenovo_krea2.safetensors

v1.0 Krea2

v1.0 Qwen2.1

v1.0 Ideogram4

v1.0 Anima_base1

v0.5 Anima

v1.0 ZImageBase

v1.0 Flux.Klein 9b Base

v1.0 Z-Image Turbo

v1.0 Flux2

v1.0 Chroma

v1.0 Qwen

v1.0 Wan

v1.0 Flux

Files with SHA256: 6967840059e54d0ca6771ae00acdc5979266f761175a8c9380f23f1a3723e9a2