ทวีตที่ปักหมุด
WisePaper × GPUhub
Building the Vibe Research ecosystem.
Structure your ideas. @wispaper
Run your models. @hub_gpu
Ship faster.
#VibeResearch
#AIResearch
#MachineLearning
#LLM
English
GPUHub
54 posts
GPUHub
@hub_gpu
On-demand cloud GPUs for AI/ML. Launch GPU servers in seconds. Flexible, scalable, pay-as-you-go.
Singapore เข้าร่วม Eylül 2022
13 กำลังติดตาม20 ผู้ติดตาม
GPUHub รีทวีตแล้ว
The full Gemma 4 family is now fine-tunable on your Mac! 🍏✨
You can now fine-tune the entire family locally with mlx-tune:
✅ E2B & E4B (Text, Vision, Audio)
✅ 26B-A4B MoE & 31B Dense (Text, Vision)
The best part? No complex routing. One unified FastVisionModel API handles it all. Want to train the built-in Conformer for ASR or Audio QA? Just flip the boolean flags and hit train.
5 complete example scripts are up in the repo! 👇
github.com/ARahim3/mlx-tu…
@GoogleDeepMind @googledevs @GoogleResearch @GoogleAI @awnihannun
English
GPUHub รีทวีตแล้ว
Most “AI case studies” I see feel like marketing slides.
The ones I actually care about are the simple, honest ones:
– here’s what we tried
– here’s how long it took
– here’s how much VRAM and money it actually used
@hub_gpu is starting to collect stories in that direction — more “real experiments”, less buzzwords. If you’re into that kind of thing, worth keeping an eye on 👇
gpuhub.com/case-studies
#MachineLearning #CloudGPU #MLOps
English
GPUHub รีทวีตแล้ว
I trained a large-scale image generation model using 8× RTX 5090 GPUs on GPUHub (@hub_gpu ) — and it was surprisingly smooth.
Here’s what happened 👇
1/3
#AI #GenerativeAI #StableDiffusion #ComfyUI #CloudGPU #MachineLearning #DeepLearning #AIArt #GPU #Tech #MLOps #BigData
English
GPUHub รีทวีตแล้ว
This is what my “ML lab” looks like now:
– modest machine at home
– rent a 24GB GPU only when I actually need it
– run YOLO/SDXL/LLM experiments end‑to‑end, then shut it down
Instead of a 4090 in my room, I get a GPU I can turn on/off like this 👇
English
GPUHub รีทวีตแล้ว
Different GPU options I’ve used:
– Colab/Kaggle → great for demos, sessions/timeouts get in the way for multi‑hour training
– RunPod/Vast → lots of raw power, but node quality/configs vary, you need to babysit jobs
– local GPU → nice latency, but you pay in upfront cost + maintenance
For most of my workloads (YOLO on non‑toy data, SDXL, 7B LoRA), the best trade‑off so far has been:
– rent a 24–32GB GPU
– treat it like a lab bench (spin up → experiment → shut down)
I’ve been using GPUhub @hub_gpu for that pattern:
gpuhub.com/?utm_source=ze…
English
GPUHub รีทวีตแล้ว
Deployed on GPUHub @hub_gpu
(Singapore-A | G067-R | RTX PRO 6000 — ~$0.91/hr)
Setup was simple, & within minutes I had a powerful environment ready.
Built a full ComfyUI pipeline with:
✔️ Stable Diffusion
✔️ ControlNet (structured outputs)
✔️ InstantID (consistent faces)
2/3
English
GPUHub รีทวีตแล้ว
I started building generative AI workflows locally… but quickly hit GPU limits 😅
Slow inference, memory issues, constant crashes.
So I moved everything to the cloud.
1/2
#AI #GenerativeAI #StableDiffusion #ComfyUI #CloudGPU #MachineLearning #DeepLearning #AIArt #GPU #Tech
GIF
English
Can AI understand code from screenshots? Yes. We tried it with Qwen2-VL-2B.
reddit.com/r/learnmachine…
English
Trained YOLOv8 on VisDrone with an RTX 5090. No instability. No wasted time. Dedicated GPUs hit different. #AI #MachineLearning #YOLOv8 #GPU #DeepLearning
reddit.com/r/learnmachine…
English
GPUHub รีทวีตแล้ว
Yaay! 🎉 Just noticed the repo crossed 800+ stars and 50+ forks.
It is honestly the best feeling to see a personal project you built for your own workflow actually helping out so many other developers.
Huge thanks to everyone testing it and contributing!
Abdur Rahim@_ARahim_
Big update! 🚨 The project has grown so fast (nearing 5k downloads!) that folks started thinking it was an official Unsloth product, even with a disclaimer on the repo. The Unsloth team was super supportive at launch, but seeing the community's confusion, they reached out. After a great chat, I’m officially renaming the project to mlx-tune to keep things clear! The vision remains exactly the same: bringing that seamless, Unsloth-like fine-tuning experience to Mac users. Same code, just a shiny new name. Update your pipelines: pip install mlx-tune
English
GPUHub รีทวีตแล้ว
Trained YOLOv8 on the VisDrone dataset (dense aerial scenes) & ran the full workflow end-to-end training, validation, inference, & export.
100 epochs finished in ~1.1h with solid results (mAP50 ≈ 0.41).
Handled crowded scenes surprisingly well.
Great experience, TY @hub_gpu 👏
English
GPUHub รีทวีตแล้ว
FREE GPU ACCESS — LIVE DROP ⚡
I’m releasing 10 GPUhub voucher codes right now 👇
Use them fast — once they’re gone, they’re gone.
💻 RTX 4090 / A800 / GPU clusters
⚡ No setup. Instant access
⚡ Perfect for AI, ML, LLM training
🎁 First come, first served.
English
github.com/aiming-lab/Aut…
I knew this was coming. Autonomous research agents are the next big thing. 🙌
English
GPUHub รีทวีตแล้ว
Someone shared their experience running RTX 5090 workloads on GPUHub Elastic Deployment in r/learnmachinelearning. Worth checking out the pros and observations.
reddit.com/r/learnmachine…
English
GPUHub รีทวีตแล้ว
More GPUhub @hub_gpu credits coming 👀
If you're interested in AI infrastructure and GPU compute:
Like and share this post.
Comment:
"drop AI gpu"
I will send another voucher.
Let's build AI together 🚀
#AI #MachineLearning #GPU
#stabledifusion #Students #learningAI #ComfyUI
English
GPUHub รีทวีตแล้ว
Have a look at GPUHub (@hub_gpu), been using them for 2+ months, their Blackwell GPUs are already at a good price, they also provide coupons (up to 45%), and you can also ask for vouchers in the support.
I just noticed that they already have something for students here: docs.gpuhub.com/promotions/stu…
English
GPUHub รีทวีตแล้ว
🚀 I'm introducing GPUhub a GPU platform that AI developers should know about.
To help people try it, I'm sharing $3 free GPU credits to test the platform.
You can experiment with real AI GPUs before renting.
1️⃣ Follow @hub_gpu Like and retweet this post
2️⃣ Create your personal account on gpuhub.com
3️⃣ Claim limited code here 👉 #redeem-vouchers" target="_blank" rel="nofollow noopener">docs.gpuhub.com/promotions/cou…
English