dispatch style h3 — modelo de IA MINIMAX H3

This is a port of my dispatch dataset from LTX to H3. It's trained on the fl2va model. All samples are t2v on 20 steps. Right now only 3 characters are trained, but I want to add in all of them. I wanted to do this so I could see if H3 can do multiple characters in 1 lora like Wan/LTX does. So far the answer is yes it can, but the voices are strange. Consider this like a proof of concept beta version. The voices are trained, along with music from the game. But when you have multiple characters in 1 scene speaking, their voices tend to mix. Especially invisi and blonde blazer in the same scene. The voices also are like 90% there, but still not super accurate. I also have seen this same problem when using the ref base model. It doesn't like to have multiple of the same gender voice in the same scene. I will keep training and tinkering to see if I can find a solution or not. Style Trigger: "Stylized 3D cel-shaded comic style." Invisigal: char_invisi, a woman with short dark hair accented by a purple streak, wearing a purple jacket over a dark top and distressed black jeans Blonde Blazer: char_bb is a woman with long blonde hair and a blue mask, wearing a blue and yellow superhero suit with a yellow cape and a red diamond-shaped gem on her chest. (no costume, powerless) char_bb has long, wavy dark brown hair and wears a strapless evening dress with long dark blue opera gloves Robert Robertson: char_rr has short brown hair and wears a light blue button-down shirt with the sleeves rolled up to his elbows and dark trousers with a beard with stubble.

Palavras-gatilho

Stylized 3D cel-shaded comic style.

618 downloads · 64 curtidas no Civitai

SHA256

C36C884810319B19ED87566D49B3D0CA47730A2D04F92C22B8B6CDEA1E7EAF10

ID de modelo ModelTrends.ai

#27052

Denunciar conteúdo

Renders de exemplo

Prompts e configurações compartilhados por quem fez estes renders no Civitai. Escolha um render para ver como foi feito.

Obter este prompt

Render de tazmannner379 no Civitai

Denunciar

Configurações

Tamanho

1920x1088

Prompt

integrated_multimodal_description: [Shot 1] Stylized 3D cel-shaded comic style. a medium close-up at a nighttime park frames char_invisi sitting on a swing, her body angled slightly to the side as she looks toward the camera with a weary expression. She has short, dark, choppy hair and wears a vibrant pink open jacket over a black crop top, with a small tan bandage on her right cheek and a round nose ring. She holds a lit cigarette between the fingers of her right hand, smoke curling upward. Behind her the park is dimly lit, a streetlamp casting a pool of light on a distant figure in a long coat walking a small dog along the path. Dark trees, a chain-link fence and the silhouettes of buildings stand under a dark sky with a few stars. The streetlamp throws a rim light along char_invisi's hair and jacket while the foreground stays in soft shadow and her silver round nose ring is slightly illuminated. The camera holds a static shot on char_invisi as she listens. Off screen and out of frame, char_rr — a man with short brown hair and light skin, wearing a light blue collared shirt — speaks to her through the radio in her earpiece. char_rr, speaking in an off-screen voice filtered through a radio (S1), says: <d>[English] Well... what about Wahn 2.5?</d> char_invisi's lips remain closed as the radio voice comes through. She takes a slow drag and exhales a cloud of smoke, then char_invisi, with a low, weary voice (S2), says: <d>[English] You don't get it! Wahn... LTX... they're old news. Stick with H3</d> She closes her eyes and flicks her cigarette back over her shoulder without looking, the lit butt tumbling away behind her in a small arc of orange sparks and vanishing into the dark. [Shot 2] Finally she stops speaking and pauses for a moment and sighs and then reaches up with both hands and grips the swing chains on either side of her, the chains creaking as they take her weight, and lets her head drop as she looks down. In the background the figure in the coat and their dog walk on and disappear off the edge of the frame, leaving the path empty as the smoke drifts and slowly thins. overall_soundscape: Quiet nighttime park ambience with faint crickets, a distant hum of traffic and a light breeze moving through the trees. A burst of radio static crackles under char_rr's filtered voice, and a soft exhale of breath carries under her words. After she finishes speaking, the swing chains creak as she grips them, small dog paws and footsteps recede along the path into the distance, and the crickets and breeze hold alone through the final quiet moment. non_diegetic_music: A sparse ambient pad at a slow tempo with a faint low drone underneath, holding quietly through the shot, then swelling into a low dramatic synth sting as she flicks the cigarette away, sustaining and slowly fading through the final quiet beat.

Render de tazmannner379 no Civitai

Denunciar

Configurações

Tamanho

1376x768

Prompt

integrated_multimodal_description: [Shot 1] 3D cel-shaded comic style, a medium shot frames char_rr sitting at his cubicle typing at his computer, seen from a three-quarter angle. He has short brown hair and light skin, wears a light blue short-sleeved collared shirt with a red logo on the chest and a large white headset with a black microphone boom, small rectangular buttons in red, green and blue on the earcup. Grey cubicle partitions, a ceiling vent and a window with closed blinds fill the background, and the monitor glow lights his face. The camera holds a static shot as his fingers move across the keyboard, the screen light flickering across his face. A red error box pops up on the screen with a sharp alert chime, and his hands stop. char_rr narrows his eyes, leans in toward the monitor, and reads it. char_rr, with a low, baffled voice (S1), says: <d>[English] What the hell is a Numpy...? And why is it out of date?</d> He sits back and rubs his temple as the error box blinks on the screen. overall_soundscape: A quiet office room tone with a faint ceiling-vent hum and distant keyboard clicks from other desks. Rapid key-tapping fills the first beat, then a sharp digital error chime cuts in and the typing stops, leaving a soft creak of his chair as he leans forward. non_diegetic_music: N/A

Modelos semelhantes

Modelos de IA para geração de imagem, vídeo e texto

dispatch style h3 é um LoRA para a família MiniMax H3 listado no ModelTrends.ai, um catálogo somente leitura de modelos de IA de código aberto para imagem, vídeo e texto. Compare-o com outros modelos MiniMax H3, veja a pontuação de calor para acompanhar a tendência e abra-o no PirateDiffusion ou BitVector para experimentar.

Mais modelos MiniMax H3 Mais LoRAs MiniMax H3 video 3d gaming comics artstyles style

Navegar por família de modelos