DoRA Weighting Note: SDXL DoRA weighting begins at 0.0. Illustrious-Dora weighting starts at 0.0, increase with 0.01 increments.
This series will attempt to be complete ai car garage. Including car of differents years models and makes. Also more attention to automive photography.
Automotive Photography
Primary concepts: car- porsche- lamborghini- parked- supercar- red- cars- lamb- sportcar- white- magazine- view- blue- photography- automotive
Photography concepts: 3 5 mm f 1. 4 photography, 4 k editorial photograph, 50 mm lens photography, 85mm photography, 8k studio photography, an ultra realistic 8k octa photo, auction catalogue photo, auto photography, automotive photography, award - winning 4 k photograph, car commercial photograph, car photography, cinematic lighting photography, color aerial photo drone, detailed sharp photo, drone photography, golden hour photography, lamborghini aventador photoshoot, low angle photography, motor sport photography, motorsports photography, photo realistic 4k, photo taken with sony a7r camera, photography shot at blue hour, red background photorealistic, shift photography, today's featured photography 4k, top down photo, top down photo at 45 degrees, uhd hyperdetailed photography, ultra realistic 8k octa photo, ultra realistic 8k octan photo, vehicle photography, with photorealistic lighting, gas station photography
Muscle Car alternate concepts:
1 9 6 5 lightning, 1 9 6 6, 1 9 7 2, 1 9 7 3, 1966, 1970, 1972, 1973, 3 ss max, 4 0 years, 4 0 years old, 40 years old, 5 0 years old, 50 years old, 70's, 80s red sports car, a black and gold pontiac car with gold rims, a red mustang mustang sitting on a black background,a white mustang with black stripes parked in a driveway, auto photography, badass batmobile car design, based on bumblebee, batmobile, blue chrome top hippo body, blue flames surrounding, bumblebee, car commercial photograph, car garage, car trading game, classic beauty, classic car, classic car magazine, classic cars, classic chrome, cliffjumper, close - up on detailed, cobra convertible, convertible, corvette c2 1969, craigslist photo, crimson - black color scheme, ctane 3 d rendered, danger lurking in the night, dark angel, dark hooded wraith, deep dish wheels, depth of field : - 2, detailed – wheels, detailed detailed, detailed detailed detailed, detailed features, detailed wheels, detailed-wheels, epic 7 0 mm lens shot, extremely clean lines, fiberglass, fine details. Red, fine fiberglass, ford, from the 7 0 s, full body detailed, full detailed, full view of a car, full view of a sport car, fully detailed, Gertrude, green 1967 ford mustang gt, green neon details, gt40, high body detail, high detailed deep, dish wheels, high detailed wheels, high-body detail, highly detailed -, highly detailed beautiful, highly detailed car, in an underground parking garage, insanely detailed c 10.0, iroc, knight rider, many exotic high end features, mopar, mustang, realistic cars, restomod, restored color, retro 1 9 7 0 s Kodachrome, retro style, saleen s7, set inside of parking garage, softglow, speedster, sportcar, sports car, sportscar, strong blue rimlit, style of hydro74, super car, supercar, the blue panther, the car is corvette c2 1969, the last v 8 interceptor, ventail, vintage car, vintage cars, white lightning, with cool headlights, yellow lightning
Description
vehicle focus, ,Aston Martin Valkyrie AMR Pro , is the focus of this model, primary color:red type:hyper car, small dataset, vehicle 44 of series.
FAQ
Comments (4)
Invoke AI
Server Error (3)
KeyError: 'lora_up.weight'
Model does not work on Invoke AI'
Lib\site-packages\invokeai\backend\patches\layers\dora_layer.py", line 34, in fr
om_state_dict_values
up=values["lora_up.weight"],
~~~~~~^^^^^^^^^^^^^^^^^^
KeyError: 'lora_up.weight'
I'll take a look. The inital issue with A1111/Forge/Comfy was that some DoRa weight required 0.0. Not the default of 1. Even though some DoRa work with a weight of 1... DoRa are more like finetunes than LoRa. I assume layers trained play a role.
It appears Invoke is not able to determine LyCoris vs LyCoris+DoRa. There is a bug open on Github for this problem.
@RokuAI thx