Mike Hong

I build AI applications for B2B. Founder at @evioslab. Partner @chilantro.

[Mike Hong @MichaelJHong]

@_aj 👏

[Mike Hong @MichaelJHong]

TLDR: Find the right timing. Connect with her.

Find her in the right emotional, mental and lovable state. Tell her how excited you are about a topic, in this case, not the Mac Pro, but the possibilities of AI and how you believe it will impact her in {topic_she_is_interested_in}.

If she is smart, she will convince you not to buy $15k Mac Pro.

[Mike Hong @MichaelJHong]

How to convince your wife in the middle of cold winter why "we" should buy 2 x Nvidia 4090s:

2 Nvidia 4090s x 400 Watts = 800 Watts x 3600 seconds (per hour)

= 2,880,000 Watts-seconds / 1054

= 2,739 BTUs per hour

We can keep our heater off at our study :-)

[Mike Hong @MichaelJHong]

Someone told me this is the new definition of “Open-Source” today.

[Mike Hong @MichaelJHong]

@dotinvoke Replicate is great. Trying to lower latency & cost & cold start.

[Mike Hong @MichaelJHong]

Any recommendations for serverless GPUs?

We’ve tried -

Banana
BaseTen
Replicate
Inferless
Cloudflare
Modal

We need scale & ease of use…

[Taelin @VictorTaelin]

RELEASE DAY

After almost 10 years of hard work, tireless research, and a dive deep into the kernels of computer science, I finally realized a dream: running a high-level language on GPUs. And I'm giving it to the world!

Bend compiles modern programming features, including:

- Lambdas with full closure support

- Unrestricted recursion and loops

- Fast object allocations of all kinds

- Folds, ADTs, continuations and much more

To HVM2, a new runtime capable of spreading that workload across 1000's of cores, in a thread-safe, low-overhead fashion. As a result, we finally have a true high-level language that runs natively on GPUs!

Here's a quick demo:

[Mike Hong @MichaelJHong]

To fine-tune or not to fine-tune, that is the question:

Whether 'tis nobler to prove thy worth, before tweaking and adjusting, or to take a model as it is, and risk the slings and arrows of outrageous errors...

To prove, to test, to validate, and then to fine-tune, ay, there's the rub! For in the pursuit of optimization, one must first demonstrate the need, lest thou fall prey to overfitting's snare.

Thus, let us not fine-tune for fine-tuning's sake, but rather, let us prove our case, and then, with careful steps, adjust and refine, that our models may serve us well, and not lead us astray.

[Mike Hong @MichaelJHong]

@SeanMollet Got it. Thanks for your insight.

[Mike Hong @MichaelJHong]

@SeanMollet Would you mind elaborating?

[Mike Hong @MichaelJHong]

@sirshibaninja Not yet.

[Mike Hong @MichaelJHong]

@catethos Never tried - will check it out.

[Mike Hong @MichaelJHong]

@dim48822 2

[Mike Hong @MichaelJHong]

Why don’t we integrate the backdrop technique directly into the chunker?

Anyone built something like this?

[Mike Hong @MichaelJHong]

Then Jim Cramer came along…

THE END.

[Mike Hong @MichaelJHong]

Once upon a time…

GPU was feeling a bit fragmented & needed some parallel processing, so it went to Therapy Processing Unit (TPU) for tensor training.

Then it met IPU at therapy and was like… whoa, I thought I live in the Matrix…

Feeling a bit less fragmented, GPU called his homie FPGA.

[Mike Hong @MichaelJHong]

HPU always talked about how he always knew HPUs that knew QPUs that knew RDUs.

[Mike Hong @MichaelJHong]

2024….

January

🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🤦‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️😴

February

🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🤷‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️💤

March
🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🏃‍♂️🙏🏃‍♂️

[Mike Hong @MichaelJHong]

@larstr Nice! Do you mind sharing the rest of your setup?

[Mike Hong @MichaelJHong]

@Clary_aiaiai @Momo_aiaiai Say yaaas @Clary_aiaiai 🤭