No, you can run “cog predict” and test on your cloud GPU before pushing.
COMPUTING
-
GPU Computing Solutions for AI Model Training
By
–
Not a stupid question at all. I currently use a VM I have access to via Replicate. But before that I used lambda labs: https://
replicate.com/docs/guides/ge
t-a-gpu-machine
… I’ve done some cpu models on Mac. cc @anotherjesse -
Tech Earnings Season: Netflix, IBM Shine Amid AI Wave
By
–
The TL Podcast for January 29th: The thick of earnings season Netflix and IBM shone while Intel and Tesla were duds as prime time for earnings season gets underway, and software and semiconductors ride a wave of excitement for AI in 2024. https://
thetechnologyletter.com/the-posts/the-
tl-podcast-for-january-29th-the-thick-of-earnings-season
… Or catch it on -
Electronics Manufacturers Beat Expectations: Signal Strong Tech Demand
By
–
Here’s a good sign for global economic trends: two contract electronics manufacturers, Celestica and Sanmina, companies that manufacture things for other companies for a fee, both reported earnings results for the December-ending quarter that were higher than expected. Sell-side
-
Code Llama 70B Release: New High-Performance Code Generation Model
By
–
Today we’re releasing Code Llama 70B: a new, more performant version of our LLM for code generation — available under the same license as previous Code Llama models. Download the models https://
bit.ly/3Oil6bQ
• CodeLlama-70B
• CodeLlama-70B-Python
• CodeLlama-70B-Instruct -
Assessing and Optimizing Your Cloud Carbon Footprint
By
–
Curious about the carbon footprint of your analytical work in the cloud? Save the date on February 13 for this LIVE discussion to learn how you can assess your carbon footprint in the cloud and ways to optimize it.
-
Google Research Advances Brain Neural Mapping Understanding
By
–
Google Research is making exciting advances on understanding how our brains work & how we think. Learn about our efforts to map the brain’s neural connections in this Research@ NYC lighting talk hosted by Viren Jain, Sr Staff Research Scientist at Google →https://t.co/oaErChf550 pic.twitter.com/aYJS2bZucW
— Google AI (@GoogleAI) 30 janvier 2024Google Research is making exciting advances on understanding how our brains work & how we think. Learn about our efforts to map the brain’s neural connections in this Research@ NYC lighting talk hosted by Viren Jain, Sr Staff Research Scientist at Google →
https://
goo.gle/3SbSClk -
JAX vs PyTorch: TPU, GPU, CPU Performance Comparison
By
–
Knowing all the development history, I'd say — JAX+XLA will be better on TPUs and slightly better on bf16 compatibility.
PyTorch will be better on NVIDIA and AMD GPUs, server-class and desktop-class CPUs, fp16 compat, vastly better on dynamic-shaped workloads.
They'll probably -
JAX TPU Performance Limits Hardware Portability Claims
By
–
would you draw a wider conclusion such as "hardware portability"?
I read the paper and concluded "JAX + TPU is unbeatable". It's hard to conclude anything more general.
The paper doesn't cover any other set of hardware — AMD, Trainium, GraphCore, etc. -
AI Models Scale Exponentially: Trillions Parameters Nine Orders Magnitude
By
–
In just a few years, cutting-edge models have gone from using millions of parameters to trillions. The amount of computation used to train the largest AI models has increased by nine orders of magnitude. Here's what we should do next: