Filter
Exclude
Time range
-
Near
5 Nov 2025
🌙 We’re diving into Echoes of the End! Let’s experience this story together.🌙🫧Sponsored by Deep Silver🫧 🚨Live now: twitch.tv/soulessiren/ #EchoesOfTheEnd #StoryStream #DeepSilverCreator
3
70
19 Oct 2025
Replying to @veefriends
„SuperReader Series“ or „Spectacular Series“ which are both just another way of saying „VeeFriends Comics“ and that they are on the website now to read! 🤝 Or it’s „StoryStream Series“ and you are telling me @djcoffman drawing streams are coming back! Either way, it’s something about the comics!
3
7
504
Hey folks, did you miss any of our preseason coverage?Maybe all of it? Well good news: all 13 of our MAC team previews are now under one storystream! Check in with your favorite teams now before kickoff: hustlebelt.com/mac-football/…
2
7
1,430
Kitten's Story Three Title: The Rewrite War It started with a rumble in the Story Vault. Kitten, Head Keeper of the Written Worlds, was reorganizing the Time Kush collection, still faintly smelling of 4/20 and sage, when the lights flickered, the air shimmered, and her cat-shaped compass began spinning like it had opinions. “That’s never good.” she muttered, sliding her detective coat over her shoulders. Within seconds, Kam burst in, out of breath, covered in ghost confetti. “Something’s wrong. The Ghost Parade just reversed. The spirits are marching backward, and Doom’s leading them with a ukulele.” Doom appeared upside-down, playing said ukulele. “They asked me to cover Radiohead.” “That doesn’t explain why Vinyl Vendetta is now a ballet.” Markus said, storming in. “My noir is being rewritten into interpretive dance.” AV stumbled through the door next, holding a smoking mirror. “This was the one you shattered in the Soul Reflection Theater, right Kitten? Because it just showed me doing taxes and being boring. I think it’s trying to rewrite my personality.” Kitten froze. “No one rewrites my stories.” she said flatly. “They're complete. Weird and perfect just the way they are.” In the Brainstorm Chamber, the team surrounded the Grand Quill, a floating relic that glowed whenever anything tried to mess with canon. It pulsed violently. “Someone is going back into our finished stories.” Kitten explained, eyes sharp. “They’re not reviving old plots. They’re rewriting the endings. Making them... tame. Predictable. Sensible.” Everyone gasped in their own weird way. “That means....” Kam started. “—no cursed dance plagues.” Markus said. “—no ghost-cheerleaders.” AV added. “—no Time Kush.” Doom whispered dramatically. “We’d all still be sober and stuck in 2024.” Kitten raised her pen like a sword. “We’re dealing with a Rewriter. The worst kind of villain. Polite. Organized. Obsessed with closure. And worst of all...” “...probably wearing beige.” AV muttered. They dove through the Storystream, hopping from one completed story to another to protect them: • In Steampunk Fate World, they stopped the Rewriter from turning Kitten into a background character named “Helper Girl.” • In Kam’s Ghost Parade, they smacked down a rewrite where the ghosts passed on peacefully without Kam singing show tunes. • In Time Kush, they blocked a clean, linear timeline edit and hotboxed the logic loops until they giggled back into chaos. • In Silent Shadows: Vinyl Vendetta, Doom dropkicked a ballet choreographer off a rooftop in slow motion. “This noir doesn't pirouette.” he growled. Finally, in the Mirror Soul Theater, they cornered the Rewriter. She wore beige. Carried a clipboard. And had notes. “I’m simply improving things.” she said sweetly. “Your stories are... messy. Why not give AV a job in accounting? Let Kam settle down. Make Kitten less confrontational.” AV dry-heaved. Kam backed away. Doom lit a fog machine on instinct. Kitten stepped forward. “I finished those stories. Not perfectly. Not neatly. But authentically. You want safe? Try a cookbook. You want weird, magical, hilarious mayhem that makes people feel something? That’s us. Now back off!” Kitten jabbed her pen into the air, and the power of stories exploded from her. The Rewriter shrieked, pixelated into feedback, and disappeared. Back in the Nexus, the gang collapsed onto beanbags, sipping cold drinks and watching their stories pulse safely on the wall. “No edits.” Kitten declared. “No regrets.” said Markus. “No dance recitals.” Doom added. Kam grinned. “Just one more round of ghost karaoke?” AV raised his glass. “To the chaos we meant to create.” Kitten smiled at her gang. “They weren’t mistakes. They were stories. Ours. And if anyone ever tries to rewrite us again…” She clicked her pen with a smirk. “We write back.” The End. (...unless someone touches the steampunk vines again, AV.)
1
2
84
🗡️✨ STREAM START – Final Fantasy XVI ✨🗡️ So, wir legen los – JETZT LIVE! Bevor wir uns dem großen leuchtenden Kristall stellen, gilt heute erstmal: Finde deinen eigenen Chocobo 🐥 Denn was ist ein Held ohne Reittier mit Stil? Kommt vorbei und begleitet mich auf diesem nächsten Abschnitt unserer Reise durch Valisthea. Ich freu mich auf euch! #FinalFantasyXVI #FF16 #TwitchDE #VtuberDE #ChocoboQuest #StoryStream #JRPG #StreamerLife #JetztLive #GERVtuber
1
2
5
216
🖤 The dojo opens TONIGHT! 🖤 ⏰ 6PM PST / 9PM EST Illuna Tales and Art are on deck. Come thru, say hi, emote, vibe, chill. .w. You never know what I got in store for the clan tonight~ 🔴 twitch.tv/MizukiArts #IllunaClan #ArtDojo #StoryStream #TwitchStreamer #SonicFanfic #CreativeStream #DojoLive #ShadowTheHedgehog #AnimeVibes #MizukiArts

ALT Fearless Year Of Shadow Year Of Shadow GIF

1
2
64
24 Jun 2025
Clair Obscur: Expedition 33 – I Still Can’t Believe That Happened… 😱🎨 #VTuber #ClairObscur #Expedition33 #IndieGame #StoryStream #mindblown Tomorrow 4pm UTC 2 (or so) youtube.com/watch?v=AyjvjOzA…
2
89
Wie jeden Samstag fangen wir heute mit ein paar Retro Jump'n Runs an, bevor es dann den wahrscheinlich letzten Storystream von Final Fantasy 10 geben wird. Hier geht's lang: twitch.tv/dervagrant
1
4
99
腾讯出的一个多模态长故事生成模型:SEED-Story 1、可以基于图片和文本,生成丰富、叙事连贯及风格一致的图片长故事 2、同时还发布了 StoryStream,专门为训练和基准测试多模态故事生成设计的数据集 3、给定相同的图像但不同的开头,SEED-Story可以生成不同的故事 github.com/TencentARC/SEED-S…
5
158
12 Jul 2024
腾讯出的一个多模态长故事生成模型:SEED-Story 1、可以基于图片和文本,生成丰富、叙事连贯及风格一致的图片长故事 2、同时还发布了 StoryStream,专门为训练和基准测试多模态故事生成设计的数据集 3、给定相同的图像但不同的开头,SEED-Story可以生成不同的故事 github:github.com/TencentARC/SEED-S… 论文:arxiv.org/abs/2407.08683
54
158
13,328
12 Jul 2024
SEED-Story TL;DR: SEED-Story 是一款能够生成包含丰富、连贯的叙事文本以及与文本中的角色和风格保持一致的图像的多模态长篇故事的 MLLM。同时,作者还发布了 StoryStream 数据集,用于构建这一模型。 TencentARC 论文地址 huggingface.co/papers/2407.0… 训练流程: -> 预训练一个基于 SD-XL de-tokenizer,通过接收预训练 ViT 的特征作为输入来重建图像。 ->> 采样一个随机长度的交错图像-文本序列,并通过在可学习查询的输出隐藏状态与目标图像的 ViT 特征之间执行下一个词预测和图像特征回归来训练 MLLM。 ->>> 从 MLLM 回归的图像特征被输入到 de-tokenizer 中,用于调整 SD-XL,增强生成图像中角色和风格的一致性。 代码已在 Github 开源 github.com/TencentARC/SEED-S… SEED-Story 模型 -> SEED-X-pretrained: 预训练基础模型 -> SEED-Story-George: 使用 StoryStream 数据集指令微调的 MLLM -> Detokenizer-George: 使用 StoryStream 数据集指令微调的 De-Tokenizer 模型地址 huggingface.co/TencentARC/SE… StoryStream 数据集 数据集源自流行的卡通系列,包括详尽的叙事文集和高分辨率图像的综合集合,它被设计用来支持长篇故事序列的创建。共三个部分: -> Curious George -> Rabbids Invasion -> The Land Before Time 数据集地址 huggingface.co/datasets/Tenc…
12 Jul 2024
SEED-Story Multimodal Long Story Generation with Large Language Model With the remarkable advancements in image generation and open-form text generation, the creation of interleaved image-text content has become an increasingly intriguing field. Multimodal story generation, characterized by producing narrative texts and vivid images in an interleaved manner, has emerged as a valuable and practical task with broad applications. However, this task poses significant challenges, as it necessitates the comprehension of the complex interplay between texts and images, and the ability to generate long sequences of coherent, contextually relevant texts and visuals. In this work, we propose SEED-Story, a novel method that leverages a Multimodal Large Language Model (MLLM) to generate extended multimodal stories. Our model, built upon the powerful comprehension capability of MLLM, predicts text tokens as well as visual tokens, which are subsequently processed with an adapted visual de-tokenizer to produce images with consistent characters and styles. We further propose multimodal attention sink mechanism to enable the generation of stories with up to 25 sequences (only 10 for training) in a highly efficient autoregressive manner. Additionally, we present a large-scale and high-resolution dataset named StoryStream for training our model and quantitatively evaluating the task of multimodal story generation in various aspects.
1
5
13
1,615
12 Jul 2024
SEED-Story Multimodal Long Story Generation with Large Language Model With the remarkable advancements in image generation and open-form text generation, the creation of interleaved image-text content has become an increasingly intriguing field. Multimodal story generation, characterized by producing narrative texts and vivid images in an interleaved manner, has emerged as a valuable and practical task with broad applications. However, this task poses significant challenges, as it necessitates the comprehension of the complex interplay between texts and images, and the ability to generate long sequences of coherent, contextually relevant texts and visuals. In this work, we propose SEED-Story, a novel method that leverages a Multimodal Large Language Model (MLLM) to generate extended multimodal stories. Our model, built upon the powerful comprehension capability of MLLM, predicts text tokens as well as visual tokens, which are subsequently processed with an adapted visual de-tokenizer to produce images with consistent characters and styles. We further propose multimodal attention sink mechanism to enable the generation of stories with up to 25 sequences (only 10 for training) in a highly efficient autoregressive manner. Additionally, we present a large-scale and high-resolution dataset named StoryStream for training our model and quantitatively evaluating the task of multimodal story generation in various aspects.
2
42
216
29,780
23 May 2024
Special shoutout to our runner ups: Lily and @konrad_gnat with WorldForager.com and Ayush Garg and team with StoryStream. Congratulations to all our participants! Check out some of the projects here: You Are The Hero: heroai.vercel.app/ Photo2Clay: loom.com/share/da262a7782c34… Vibes Video Search Engine: vibes.video, github.com/JeremyNixon/video… seeksequence: seeksequence.com WorldForager: worldforager.com, github.com/kon-rad/worldfora…
1
1
3
517
The Verge's StoryStream short content launched in September 2022 has worked out really well for them. I often see their short content ranking as snippet (outranking source) even though its a comment on research or an article published in some other publication.
1
6
405
Live!!! Let's ride some Rollercoasters together! Also, we are only 4 away from my follower goal and bedtime storystream! #VTuberEN #VTuberUprising twitch.tv/kaidavindur?sr=a

2
35
I've made it through the #Bears RBs, TEs, WRs, and OL, so only one position left on the offense to review for tomorrow morning. I'm sure the comment section will be civil and respectful of all differing opinions. Storystream here to catch up: windycitygridiron.com/2024/2…
We're almost all the way through the offense in @wiltfongjr's Chicago #Bears 2024 roster turnover series, so check out his look at the offensive tackles... windycitygridiron.com/2024/2…
2
1
12
5,561
20 Nov 2023
If you're having trouble keeping track of the OpenAI situation, we've been collecting all the newest info in a handy storystream. Follow along for the latest updates, including how Harry Potter fanfic factors into all of this: theverge.com/23966325/openai…
2
23
200
59,021