Sign In

Jennifer Connelly 1990s (HUNYUAN + FLUX + SDXL)

194
3.2k
97
Type
LoRA
Stats
662
0
Reviews
Published
Feb 4, 2025
Base Model
Hunyuan Video
Training
Steps: 3,500
Epochs: 50
Usage Tips
Clip Skip: 1
Strength: 1.1
Hash
AutoV2
7C2BFF054E
Tencent Hunyuan is licensed under the Tencent Hunyuan Community License Agreement, Copyright © 2024 Tencent. All Rights Reserved. The trademark rights of “Tencent Hunyuan” are owned by Tencent or its affiliate.
Powered by Tencent Hunyuan

HUNYUAN v1.0 :

Trained on Hunyuan Video fp8 with 512x512 px with 70 photos of Jennifer Connelly between 1989-1996 with detailed GPT-4 captions. Tested on Hunyuan Video fp8 and Fast Hunyuan Video fp8 ! No keywords needed. Use around Lora strength=1.1.embedded_guidance_scale= 5.0-8.0 and flow_shift=7.0-12.0:

Positive : {Short summary of the scene e.g. Professional video of a blonde woman giving and interview on the red carpet}, {more detailed scene and background description}, {lighting description}, {camera direction e.g. panning in, panning out, zoom in etc.},<lora:jenconnelly_1990s_hunyuan_epoch50:1.1>

FLUX v1.0 :

Please Donate Buzz for FLUX Lora Training !

Trained on FLUX.1 [dev] with 99 photos of Jennifer Connelly between 1989-1996. Tested on FLUX 1.D (full) and FLUX fp8 and FLUX nf4 ! Use around strength 0.8-1.2. No keywords needed! Distilled CFG around 1-4 and CFG 1.0 (without negative prompt). Clipskip 1. Can be used for example as follows:

Positive : {Artstyle, Character and scene description in usual FLUX fashion}, <lora:Jennifer_Connelly_FLUX_v1-000049:1>

SDXL v2.0 :

Trained on Juggernaut X with 131 photos of Jennifer Connelly between 1989-1996. Tested on Juggernaut X, Juggernaut v7, RealismEngine 2, RealVisXL3, AlbedoBaseV2! Use with keyword : "jenxcnnlly" . Use around strength 1.0-1.1. CFG 5.0-7.0. Clipskip 1. Can be used for example as follows:

Positive : {Artstyle}, {Character and scene description}, jenxcnnlly, <lora:jenconnelly1990s_juggerX_xl_1_wocap_merger_71_110_150_03_04_03-jenxcnnlly:1.1>

Negative : ugly, deformed, airbrushed, photoshop, rendered, (multiple people), child

SDXL v1.0 :

Trained on 131 photos of Jennifer Connelly between 1989-1996 on SDXL 1.0 base. No keywords needed! Lora strength with prompt should be around 0.8-1.2. Following prompt structure on Juggernaut v7, realismEngine 2.0 or Juggernautv9 lightning works well:

positive : {art style}, {character descrioption and action}, <lora:jenconnelly90s_xl_1_standard_wo_cap_merger_35_55_93-025_035_04:1>

negative : photoshop, rendered, (multiple people), blurry, child