Adédọ̀tun𓂀 (@adedxtun) header
Adédọ̀tun𓂀 (@adedxtun) profile picture

Chasing Redemption

In a server near you
Joined February 2020
Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)Photo by Adédọ̀tun𓂀 (@adedxtun)
Picked up Inference Engineering because I think it's a powerful intersection of the things I enjoy and it's going to be in any and everything. I'd keep sharing the things I'm learning on this thread.
2
2
151
Adédọ̀tun𓂀 retweeted
This best summarized the ideal internal resolve of a man. Listen.
185
5,960
765
27,140
2,415,804
Chapter 3 - Hardware (Philip Kiely’s Inference Engineering) GPUs are throughput designed simply for massive parallel workloads. Model inference involves a lot of matrix multiplication, so the hardware matches the work.
1
3
122
What limits it? Personal devices have less compute, plus heat and battery constraints. Supporting many different devices is also harder.
1
9
What fits? Desktops can run larger models, especially when quantized. Phones are better suited to smaller tasks such as transcription, speech, or translation. The chapter’s conclusion is a hybrid approach: run small, quick tasks locally and send demanding work to datacenter GPUs
9
Adédọ̀tun𓂀 retweeted
As a scalper, I want to trade the setup I’m looking at without opening a sidebar to enter my position size. That’s what we’ve built on Furge. Add a long or short drawing, drag it to adjust your risk, and size your position directly on the chart. The small popover gives you the controls: amount, leverage and which exchange to trade on. Hit “Trade” and the order goes through. No extra confirmation step. From setup to execution, right on the chart.
1
3
1
8
586