Chromadera (@Chromadera) header
Chromadera (@Chromadera) profile picture

At the precipice of success Local ai Qwen-Lover AMD Rocker

Joined May 2023
Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)Photo by Chromadera (@Chromadera)
Finally convinced to get off API and try the plans directly
1
46
Google and @OfficialLoganK missed this opportunity. This model is that good. It's the best thing Google has ever released in the race to AGI. ALLOW US USE IT OUTSIDE ANTIGRAVITY.... Btw Neo>>>>Neon
🤖 Made with AI
2
52
Chromadera retweeted
Ref version of the Frieren LoRA for MiniMax H3 is out! Trained on the ref model (with @ostrisai’s ref training adapter v3, and musubi fork), so it works better with ref2va than the FL2VA One. Same character sheets + voice refs from the HF repo below. Let me know what you think! huggingface.co/tazmannner/Fr…
released my Frieren LoRA for MiniMax H3 every scene in the video is H3 + the LoRA. colors, line work, animation timing and even the bgm are from the show's style now also included: 37 character sheets, generated voice refs for 27 characters, a prompt guide and the full training write-up. My most ambitious style lora yet: huggingface.co/tazmannner/Fr…
2
4
27
1,319
Chromadera retweeted
I recommend the 3600 steps checkpoint, after this validation started to plateau and trend negatively. This is trained on the same exact parameters, including audio as the FLV2A lora. I didn't extensively test it though, and I just generated this samples and called it a day. If I find any issues or a different checkpoint is better I'll share. I wanted to get this out the door quickly so I could work on other things. Hope it's helpful!
1
1
85
previz is so stressful when working with multiple characters. here is first frames for 6+
1
1
5
252
Chromadera retweeted
Gemini 4 Carbon - Rendering a beautiful woman. Doesn't she look like that spy who had an affair with Jensen?
1
1
7
327
Chromadera retweeted
People keep asking me which subscription to get
80
42
8
1,416
86,076
Chromadera retweeted
making your own software is tough, even with AI, because it's hard to know what you really want I didn't end up sticking with this, the modes weren't the right way for me to organize my work gonna a different take that’s more to-do oriented
experimenting with a lil personal project: an AI-native macOS dock with custom modes (writing, building, etc.) to help me stay focused my agent is able to customize mode behavior and add new modes, it uses jev to classify my activity + notifications in real time & act on them
131
47
32
2,058
149,558
Chromadera retweeted
big video, small model. Rumble-2-Flash is now open weights. 40+ languages, Apache 2.0, 0.5B parameters. download it: huggingface.co/Reson8/Rumble…
1
5
2
11
332
Chromadera retweeted
bravo monsieur temkosoft c'est merite c'est le meilleur jeu de volley du monde
volleyball is back!
4
13
198
1,759
Make your own vegan hardcoded version then. Life isn't that hard
it sucks because this is really really cool but it uses ai which means it’s worthless to me. stupid dogshit llm code slop
106
Chromadera retweeted
brand exploration / nullbox
8
3
49
1,150
Training a lora... Drawbacks so far from h3 (filling in the loops I didn't specify with its own garbage). Background characters are poorly done. Internal schematics of the bus is off. Texts and inscriptions are wacky but all of this can easily be fixed with one coordinated parsing and previz for the motion hehe.
5
196
Replying to @Jalann_art
Don't get me started on how everyone who's Black or White HAVE to be represented as American. No other ethnicity matters.
9
20
3
1,120
20,046
Chromadera retweeted
idea guys still aren't bothering to do anything so that hasn't changed but devs finally cranking out all their side projects now we can know for sure all our ideas sucked
95
86
17
2,495
56,741
Chromadera retweeted
Replying to @itsmissglitch
They don't have to make the best coding model either. I'll be happy with a slightly smarter Gemma 31b but much better tool calling. Something that can drive my custom telegram harness. It's so much better to talk to than current models, frontier or otherwise
1
1
3
57
What you are lifting there is 16KG per arm. The rod itself weighs 1Kg. 4x 2.5kg + 2x 1.25kg + 1kg for the rod itself. The least you should do before laughing at regular people is know your weights.
Since y’all were asking if I could do it, I just overhead pressed 20kg in each hand for about 15 reps without struggling. But you’re telling me it’s normal for grown men to struggle with a total of 25kg? That’s just 12.5kg per arm! Are we saying most men simply don’t train their shoulders, or am I expecting too much?
2
4
260
I don't know if that tinier weight at the end is 0.5kg which means at most you are pushing 17Kg which is my initial estimate
2
56
For a sick Theo to get up and hate .. Your model must have really really dropped the ball super super bad.
6.2 Astra is here. I’ve been testing it for awhile. I think this is the worst model OpenAI has ever released. I’ll share more soon.
1
3
99
Chromadera retweeted
说个暴论,真搞不懂现在怎么还有那么多教人用 AI 的教程。 现在的 AI 早就不是前几年那种理解力有限,非得背一套提示词模板才能干活的时候了。打开软件,直接用日常语言像聊天一样把想法说清楚不就行了吗?连正常说话都要看教程学,缺的根本不是什么工具技巧,就是基本的表达和沟通能力。 真正该花时间去学的,反而是专业本身的底子。就拿写代码来说,现在的门槛根本不在怎么用 AI,而在软件工程知识。项目结构怎么拆合理,边界条件怎么兜底,还有 AI 生成的代码到底靠不靠谱、有没有暗坑,这些全得靠自己的工程能力去判断。放着正经的软件工程不学,天天研究怎么在聊天框里提问,纯属本末倒置。
55
14
9
305
19,386