Nicolay Gerold

- Give customers a no brainer: make friction crazy low (so painless to add it, now you got exposure and you can pop out when relevant when you can provide the value you can offer them their savings and advertise a real account and take your percentage of the savings), e.g. chrome extensions like Honey, or give them money back with an economical ret... See more

Detach from the code itself

Amazing engineers stay unattached to the code itself.

They were unafraid to delete and start over on code, even if they were 90% of the way in, if it meant that the end result would be better overall.

Code isn’t personal, so feedback was taken with stride.

Code isn’t perfect. Nobody cares about perfect code. They care abo... See more

4. Introducing Stable LM 3B: Bringing Sustainable, High-Performance Language Models to Smart Devices

Stability AI introduced Stable LM 3B, a high-performing language model designed for smart devices. With 3 billion parameters, it outperforms state-of-the-art 3B models and reduces operating costs and power consumption. The model enables a broader ran... See more

ETL

The part of the system I'm most proud of, and on which I spent the most effort, is the ETL process.

We had a series of shell scripts for each data source we ingested (there were many), which would pull the data and put it in an s3 bucket.

Then, early in the morning, a cron job would spin up an EC2 instance, which would pull in the latest ETL code... See more

Model Explorer is a powerful graph visualization tool that helps one understand, debug, and optimize ML models. It specializes in visualizing large graphs in an intuitive, hierarchical format, but works well for smaller models as well.

Graph visualization plays a pivotal role in the machine learning (ML) development process. Visual representations o... See more

AI Template & SPR Library

Featuring advanced prompts and SPRs

🟢 Website

🔵 LinkedIn

🔴 Patreon

⚪ Discord

Prompt Engineering

Advanced GPTs

Template Library

Using feature flags

It's always a good idea to put new features behind a feature flag.

This contributes to a rollout strategy that can surface user feedback in a much more contained and safe environment.

When thinking about how to rollout a new feature to users, first think about your peers. Many of the Resend team members also use Resend on their ow... See more

The Mac Studio is an absolute monster for inferencing; but there are a couple of caveats.

• Its slower, pound for pound, than a 4090 when dealing with models the 4090 can fit in its VRAM. So a 13b model on the 4090 is almost twice as fast as it running on the M2.


• The M1 Ultra Mac Studio with 128GB costs far less ($3700 or so) and the inference speed is

... See more