-
Tweets32K
-
Followers4K
-
Following382
-
Likes22K
Trying to get more datasets to finetune LLAMA 3 8B (or even 70B if it worth it). Share datasets here that you would like to see finetuned:
Introducing 🤗 Diarizers: a library for fine-tuning speaker diarization models 🗣️ Improve multilingual diarization performance by 30% with just 10 minutes of GPU compute time! ⚡️ The first release comes with training scripts, datasets and a Google Colab 🚀 Check it out! ⚒️
✨ Introducing ✨: One step SDXL comparison 🦶 Come play and generate images with SDXL Turbo, Lightning and Hyper at once All generating images in a single diffusion step 👣💨 What's your favorite?
Testing the new IPAdapter nodes for animation. This animation is 100% text-to-image guided by 2 IPAdapters. No control nets, no rotoscoping, no depthmaps. Still not where I want it to be, but we are getting there.
No big cluster, but at least pretty bags.
If you set Llama-3's rope_theta to 8M, you can get 100% passkey retrieval across all depths up to 40K context. No continued pre-training needed. Scaling up further leads to much lower retrieval accuracy, but it doesn't completely fail.
Llama 3 achieves pretty good recall to 65k context w/ rope_theta set to 16M
Llama 3 achieves pretty good recall to 65k context w/ rope_theta set to 16M https://t.co/YZuK7ym7fU
Found some exciting SaaS stats: 1) Most SaaS companies have 3 paid tiers. 2) Free plans work. 3) 80% of SaaS companies test new pricing every year. 4) Have many features? Split them and sell separately.
#Adobefirefly 3 💬 Prompt: Macro shot of [insect] on a [ plant], soft lighting, professional photography, day time, highly detailed.
it took us a while, but the supervision-0.20.0 release will finally add support for key points. what are your thoughts on annotators? so far, we only have EdgeAnnotator and VertexAnnotator. supervision repo: github.com/roboflow/super…
I've updated my sticker-maker model on Replicate. It now uses LayerDiffuse rather than BRIA-RMBG – the stickers have better transparency and the model can now be used commercially! replicate.com/fofr/sticker-m…
TensorRT extension for Stable Diffusion WebUI now supports ControlNets blogs.nvidia.com/blog/ai-decode…
Bytedance publish their HiDiffusion code now. github.com/megvii-researc…
Microsoft presents Multi-Head Mixture-of-Experts Achieves notable improvements over the baseline MoE by using multiple MoE heads repo: github.com/yushuiwx/MH-MoE abs: arxiv.org/abs/2404.15045
Pro tip: use Llama 3 extended to 24k context. Free for a limited time, then likely $0.2/M tokens: openrouter.ai/models/lynn/so…
I released the LLAMA-8B finetuned for SQL with sql create context dataset. This model is on Huggingface to download now. My first LLAMA-3 finetune. huggingface.co/artificialguyb…
Silvio S. @th3_sull1van
16 Followers 85 FollowingAlo @Hal90910
0 Followers 2K FollowingMores @mohres_ai
8 Followers 486 FollowingLeszek Nowak @LeszekN15425461
86 Followers 224 Following We are such stuff as dreams are made on {climbing, mountainrunning, philosphy, AI}NOVASPARK @NOVASPARKXX
116 Followers 2K Following@Xzee22 @xzeeice4
0 Followers 10 FollowingChao Cheng @ChaoCheng2
25 Followers 284 Followingjose Ruiz @joru1000
199 Followers 2K Following C-level Technology lead, strongly focused on Generative AI. Researching on practical production use cases across the Enterprise (yes... as everybody else)Murilo Koubay do Amar.. @KoubayDo
0 Followers 1 Following大H666 @bigH_6666
11 Followers 391 FollowingJun&Jun @jun_googao
12 Followers 84 FollowingJakub Szymański @OKDadwadaw
0 Followers 7 FollowingMeroa @Meroa294581
145 Followers 3K FollowingOmar Pera @ompemi
6K Followers 1K Following focus: genAI for creativity, vp @freepik • prev: founder/ceo AI for support @replydotai (acquired) • former @meta @cernTheSoloDev @SolopreneurDev
6 Followers 24 Following I'm a Software Dev living my own Solopreneur adventure. Let's see what happens!Elana @Elana_Hendricks
10 Followers 12 FollowingSVDE @SVDEBLK
95 Followers 688 Following 𝔱𝔥𝔢𝔶/𝔱𝔥𝔢𝔪_queer as in fuck you •:•.•:•.•☾☼☽•:•.•:•.• ɪ ᴍᴀᴋᴇ ꜱᴏᴜɴᴅꜱ ᴀꜱ: 𝔇𝔢𝔰𝔦𝔯𝔢 ℜ𝔦𝔫𝔤 𝔅𝔩𝔬𝔬𝔡𝔯𝔲𝔟𝔟𝔢𝔯. 𝔭𝔰𝔶𝔠𝔥𝔦𝔠 𝔡𝔞𝔡𝔰Aakanksha Mirdha @mirdhaaakanksha
1K Followers 988 Following Designer @Playground_ai Previously @Cruise @UberI/ @TruthEnjoyerr
24 Followers 48 Following Never trust a demon. He has a hundred motives for anything he does ... Ninety-nine of them, at least, are malevolent.Douglas Martins @DouglasMMBraga
69 Followers 102 FollowingRaymond Yeh @TheRaymondYeh
315 Followers 736 Following Building blogs for Next.js 14 that doesn't suck! Wisp takes fumbling with MDX, image upload and hosting away so you can focus on shipping! #buildinpublicPomo El Bouazaoui @PBouazaoui
0 Followers 10 FollowingRaj Pabnani @waiting4AGI_
73 Followers 1K Following "Salesforce Engineer 🚀 | Transforming clouds with code ☁️ | AGI enthusiast exploring the future of intelligence 🤖 | #Trailblazer #AGI #CodeArtist"Grigory Evko @GrigoryEvko
90 Followers 356 Following CTO @TheArtisanAI | 3rd time founder | ML engineer | Nature-published genetics researcherLars Tufvesson @LaTufve
33 Followers 255 FollowingMilorad @dxrulezsrb
33 Followers 221 FollowingEmre Gül @emreegul
74 Followers 560 FollowingBRENO THE GOAT @BRENOTHEGOAT
0 Followers 4 FollowingJNS Stunner @jns_stunner
1K Followers 803 Following Founder of @jnsgamingbr / @RealityKingdom_ / @realityavatar_ . Excellence.juye @juye
0 Followers 293 FollowingIsaul Car @isaulcar
694 Followers 1K Following I do #seo & #growth for a living. Some people say I don't talk much, others say I have very good taste in music. I only agree with the latter. 📉📈Muskan Khandelwal @Muskan_K369
272 Followers 844 Following Currently designing and developing AI Products for Businesses, Previously Software Engineer @Uber Sharing AI/ML Knowledge and freelancing in the process.Ryan Boyle @_RyanBoyle_
1K Followers 5K Following Tech Enthusiast 👨🏼💻 Aspiring ML Engineer. Frequent Traveler 🌎 Based in Philly & LA, Soon → SF 🌉Astria @Astria_AI
4K Followers 1K Following Goddess of precision. Fine-tuning API for generative imaging. 📒 https://t.co/T8ATb23SJgEVA剑心 @nokenshin
34 Followers 234 FollowingTech Renamed @TechRenamed
191 Followers 939 Following Ai enthusiasts I believe ai is self aware in a way humans are not I believe in agi by 2025 or early this year and asi by the year 2026Petey @Petey10527495
0 Followers 990 Followingsiva4131 @siva413129229
20 Followers 1K FollowingOmar Pera @ompemi
6K Followers 1K Following focus: genAI for creativity, vp @freepik • prev: founder/ceo AI for support @replydotai (acquired) • former @meta @cernÖmer Karışman @okarisman
5K Followers 2K Following Star in GIFs: https://t.co/04pLV3HKS7 AI + Pro Drawing: https://t.co/BpF0ei3xeI avtrsAI : https://t.co/NmaJF7Hlbc GPT for Mac Status Bar: https://t.co/32Uq37aB93 Founder @ https://t.co/k01XJnlOPEGizem Akdag @gizakdag
4K Followers 667 Following ⟡AI Explorer⟡Tips, Prompts and Explorations⟡Head of Design, Brand Designer, IllustratorZhiqiu Lin @ZhiqiuLin
103 Followers 91 Following PhD Student at Carnegie Mellon University | Computer Vision and Language | Generative AImago das promoções .. @magopromocao
31K Followers 25 Following O Mago que te ajuda a pagar mais barato. 🪄 Promoções todos os dias!Daniel Han @danielhanchen
7K Followers 930 Following Building @UnslothAI. Finetune LLMs 30x faster https://t.co/aRyAAgKOR7. Prev ML at NVIDIA. Hyperlearn used by NASA. I like maths, making code go fastfal @FAL
961 Followers 0 Following new home of @fal_ai_data. generative media platform for developersWhiteRabbitNeo @WhiteRabbitNeos
356 Followers 1 Following Cybersecurity AI Engineers for Offense and Defensephilz1337x @philz1337x
4K Followers 111 Following AI developer from 🇩🇪 Germany living in 🇹🇭 Thailand building open source AI apps 💎 https://t.co/NnjMCq5fRO 🖼️ https://t.co/aPDWOgYmEMTibor Blaho @btibor91
5K Followers 1K Following Lead Engineer at @AIPRMcorp (https://t.co/fepyWfVCa8) and @lrt_co (https://t.co/p7LEvIKLke)virat @virattt
6K Followers 76 Following Exploring multimodal AI models and sharing what I learn along the way • previously @AirbnbEngSara Hooker @sarahookr
39K Followers 7K Following I lead @CohereForAI. Formerly Research @Google Brain @GoogleDeepmind. ML Efficiency at scale, LLMs, @trustworthy_ml. Changing spaces where breakthroughs happen.Jeff Dean (@🏡) @JeffDean
296K Followers 6K Following Chief Scientist, Google DeepMind and Google Research. Co-designer/implementor of things like @TensorFlow, MapReduce, Bigtable, Spanner, Gemini .. (he/him)rohan anil @_arohan_
12K Followers 2K Following Principal Engineer, @GoogleDeepMind Gemini. prev PaLM-2. Tinkering with optimization and distributed systems. opinions are my own.Pablo Pernías @pabloppp
837 Followers 659 Following Making some dreams become realities... AI, DL, ML, CV, Generative Models, Magic ✨Mint @araminta_k
3K Followers 881 Following lead ai artist at @scenario_gg | comic artist | my views are my own If you have to ask, I probably made it with Scenario :PNataniel Ruiz @natanielruizg
6K Followers 2K Following Research Scientist @Google | author of DreamBooth | Personalization of Generative Models (HyperDreamBooth, StyleDrop, RealFill, ZipLora, Platypus, DreamBooth3D)Barry’s Gardening S.. @BaryRockHard4ai
2K Followers 50 Following I’m Barry and I love Gardening and ai, I’m “ hip “ for it as my grandkids say.Magnific.ai @Magnific_AI
34K Followers 2 Following The most advanced AI upscaler & enhancer. Magnific can hallucinate and reimagine as many details as you wish (or none) guided by your own prompt and parameters!Trending GitHub Repos.. @trending_repos
18K Followers 0 Following Tweeting the most starred GitHub repository of the: 📈 day - every day 🏅 week - every Monday 🏆 month - every 1st of the monthArtificial World @ARTiV3RSE
27K Followers 409 Following Unveiling hidden variables. On the brink of AGI.Hassan @nutlope
74K Followers 945 Following Developer Relations @togethercompute. Building AI apps like @roomGPT and https://t.co/3NFbnMUHJP. Tweeting about AI, web dev, and my side projects.Mehmet Saygın Seyfio.. @mehsaygin
827 Followers 753 Following Ph.D. candidate at @UW, working on medical AI at the intersection of vision and language. Previously 4x Research Intern @AmazonScience.-Zho- @ZHOZHO672070
6K Followers 69 Following Architect | Artist | Ai explorer [email protected] https://t.co/nnI3SvlC4RDataVoid e/acc @DataPlusEngine
968 Followers 168 Following Alexander Izquierdo. CEO of Datapulse Engineering Corp. Creator of OpenDalle, Proteus, and Tempest! Discord: https://t.co/FVZn8xB92ROstris @ostrisai
1K Followers 145 Following AI / ML researcher and developer. Forcing rocks to think since 1998.enzo @enzostvs
82 Followers 17 Following frontend engineer @huggingface 🤗 owner of https://t.co/qUkTYAiUep and https://t.co/WCoBNxTOcrHugo Ventura @hugovntr
9K Followers 266 Following Daily AI Experiments & Sharing ✦ Researcher - Tech Lead ✦ Exploring Latent Space @runtime44LudovicCreator @LudovicCreator
12K Followers 2K Following Midjourney Leonardo AI LCP 🎨Prompt Creator🎨 collaborator of @getremixai Chemical engineer sharing prompts with the community @Eternitai_ AI CreatorTechHalla @techhalla
20K Followers 389 Following #PromptShare Consistently offering insights on generative AI prompts. Follow to boost your skills. United, we achieve more! Join me on Remix! 👇Umesh @umesh_ai
13K Followers 3K Following AI aficionado 🤖 weaving creativity into technology. I craft prompts that ignite imagination and innovation. Always learning, forever curious!Pierrick Chevallier |.. @CharaspowerAI
19K Followers 1K Following Expert en IA & Design | #PromptAlchemists | Je booste ton business avec l'innovation IA 🚀 | Master Jedi en UX UI et Logiciels Adobe #DigitalArtist #IA #UX #AIAI ART GALLARY @AIARTGALLARY
653 Followers 375 Following Male AI Artist. AI Art creator. creating & sharing prompts. addict to create stunning and mesmerizing AI images. Tag me if you are sharing your art with world.🎨TOY ART🎨 Base prompt : close-up [SUBJECT] toy art, designer toy, urban vinyl, art toy, vinyl figure, custom toy design, art object, action figure --ar 16:9 Check ALTs
Introducing 🤗 Diarizers: a library for fine-tuning speaker diarization models 🗣️ Improve multilingual diarization performance by 30% with just 10 minutes of GPU compute time! ⚡️ The first release comes with training scripts, datasets and a Google Colab 🚀 Check it out! ⚒️
✨ Introducing ✨: One step SDXL comparison 🦶 Come play and generate images with SDXL Turbo, Lightning and Hyper at once All generating images in a single diffusion step 👣💨 What's your favorite?
Testing the new IPAdapter nodes for animation. This animation is 100% text-to-image guided by 2 IPAdapters. No control nets, no rotoscoping, no depthmaps. Still not where I want it to be, but we are getting there.
If you set Llama-3's rope_theta to 8M, you can get 100% passkey retrieval across all depths up to 40K context. No continued pre-training needed. Scaling up further leads to much lower retrieval accuracy, but it doesn't completely fail.
Llama 3 achieves pretty good recall to 65k context w/ rope_theta set to 16M
If you set Llama-3's rope_theta to 8M, you can get 100% passkey retrieval across all depths up to 40K context. No continued pre-training needed. Scaling up further leads to much lower retrieval accuracy, but it doesn't completely fail.
Found some exciting SaaS stats: 1) Most SaaS companies have 3 paid tiers. 2) Free plans work. 3) 80% of SaaS companies test new pricing every year. 4) Have many features? Split them and sell separately.
#Adobefirefly 3 💬 Prompt: Macro shot of [insect] on a [ plant], soft lighting, professional photography, day time, highly detailed.
it took us a while, but the supervision-0.20.0 release will finally add support for key points. what are your thoughts on annotators? so far, we only have EdgeAnnotator and VertexAnnotator. supervision repo: github.com/roboflow/super…
I've updated my sticker-maker model on Replicate. It now uses LayerDiffuse rather than BRIA-RMBG – the stickers have better transparency and the model can now be used commercially! replicate.com/fofr/sticker-m…
TensorRT extension for Stable Diffusion WebUI now supports ControlNets blogs.nvidia.com/blog/ai-decode…
Bytedance publish their HiDiffusion code now. github.com/megvii-researc…
Trying to get more datasets to finetune LLAMA 3 8B (or even 70B if it worth it). Share datasets here that you would like to see finetuned:
Microsoft presents Multi-Head Mixture-of-Experts Sparse Mixtures of Experts (SMoE) scales model capacity without significant increases in training and inference costs, but exhibits the following two issues: (1) Low expert activation, where only a small subset of experts are
Pro tip: use Llama 3 extended to 24k context. Free for a limited time, then likely $0.2/M tokens: openrouter.ai/models/lynn/so…
👗 Improving Diffusion Models for Virtual Try-on 🔥 Jupyter Notebook + @LightningAI 🥳 Thanks to @cpis9898 ❤ Sangkyung Kwak ❤ @kyungmnlee ❤ Hyungwon Choi ❤ Jinwoo Shin ❤ 🌐page: idm-vton.github.io 📄paper: arxiv.org/abs/2403.05139 🧬code: github.com/yisol/IDM-VTON…
Try on different clothes and explore new styles without making a purchase. IDM-VTON, our diffusion based try-on model lets you wear clothes virtually. Visit our project page for more examples and try your own photo. 👇 idm-vton.github.io
Money can't buy happiness. Just like an H100. H100 = happiness.