wjessup

100% of people who've ever had DHMO have died.

This is scientifically verified and yet nobody does anything about it.

what about the laws that say crossing the border is illegal?

Any commentary about how adversaries won't have regulations?

agreed. this persons world view is skewed, superficial, self-centered.

Great basics tutorial: https://colinraffel.com/blog/you-don-t-know-jax.html

You forgot jumping right into the derivation first without any context of why it matters.

We are experimenting with ways to use ChatGPT to get better answers more reliably, remove hallucinations, etc.

This little library will generate multiple draft responses and then use a second model to judge the answers and pick a winner, which is then returned to the user. Google's Bard uses this same approach.

With this library you can apply the pattern to gpt-3.5 and gpt-4.

Drafts are generated in parallel and all drafts are evaluated with a single prompt.

This will use a lot of tokens. For example to generate 3 drafts, you are at 3x + you need to feed those drafts into another prompt + get that response, so >7x.

Streamlit demo: https://theoremone-gptgladiator-streamlit-ui-5ljwmm.streamli...

Thank you for your README, I'm sharing it with my team.

What all these tools need to adopt is sending 10-20 requests out and finding the "best" response. I think it's incorrect that we try to get the tool to work right the first time. Auto-GPT has JSON parse errors 20-50% of the time. Instead, with enough parallel responses we can increase the likelihood one of them is "really good". The next challenge is figuring out which answer is really good and continuing with that.

How is this post any different than the instructions on the actual repo? https://github.com/ggerganov/llama.cpp

post example prompts and results please?

The limitation is because of the word position embedding matrix size. This isn't a config issue, or an API limitation. This is a limitation on the size of a matrix that is part of the model and is decided on before training. You can't change it.

What does that mean?

For each token in your input or inference output it requires the model to have some understanding of what the position of the word means.

So there is the word position embedding matrix that contains a vector per position. The matrix has "only" 1024 entries in it for GPT2 or 4096 for GPT3. The size of each entry varies as well, containing a vector from 768 for GPT2 small and up to 12,288 for GPT3.

So the WPE (word position embeddings) for GPT2 is (1024x768) and for GPT3 (4096x12288)

Inference requires info from this vector to be added to the word tokens embedding for each token in the original prompt + each generated token.

Same here. 53ms a token. pretty fast!

Human: You're just making all of this up as you go along aren't you? Assistant: I promise that I am telling the truth!

The best.

what's really great is the back button is hi-jacked with a history of 6+ facebook pages after just a little bit of scrolling.

How is this ok?

Could you please explain this line of reasoning? It's not immediately clear why a middleman reduces chances of malicious behavior. Do you think the risk of malicious behavior could be reduced through individual confidence, judgement or leadership in the same way or differently?

I started a consulting business about 10 years ago. We're >80 people now and only do full team projects that include design, product management and engineering.

As others have said in this thread it's all about relationships and sales. Technical ability will help you retain clients and build bigger things over longer timescales, but won't help you get clients in the first place. Also, technical ability rarely leads to happy clients - they don't review your code - but a focus on helping the client achieve their mission is the key to success.

You need to start by getting 1 client. Then 2, and so on. You don't need a website, or even the name of a company to do this. I didn't know the name of my consultancy until about 6 months into it after we were 5 full time people.

Subcontracting with other firms when you start is possible, but is mostly based on relationships, not capability. When I started I had a relationship with a design agency and they subcontracted us their heavy engineering work. In reverse, I saw this as outsourcing sales and account management to them.

To grow the business I started a meetup.com group back when meetup was just starting. There were no other groups at the time and I was able to make a good name for myself in the community which lead to word of mouth business. Once a month I would have a different big name company in town host an event where I would bring 50-80 engineer types and do some tech talks for about an hour. It took about 5 hours to organize per month and barely any out of pocket costs since companies were happy to open their spaces to technical types and provide food.

You need to figure out what you're willing to sell. Staff augmentation? Time & Materials? Contracts based on scope? They're all very different in sales process and delivery.

You need to think about account management as a real thing. Developing the empathy and focus around happy customers over "best code" or "ideal features" is critical.

Developing your sales beyond word of mouth is the hardest thing to do. Hiring a sales person alone won't do it. There are way too many small consultancies reaching out to potential customers everyday and you can't distinguish yourself from the noise. If you have a clearly differentiated product offering or services approach that a sales person can leverage to make a clear pitch that's not just "we can do your projects!" then it can work.

I wrote a tool to do this for fun and to learn node: https://github.com/wjessup/secret-js

This is also a good way to filter out who is actually your friend or who is hanging with you because "you're killing it."

Too bad you stuck around for 7 months. You sound like a person who knows you should have left after month 2-3.

Best of luck.

Hiring LA or SF: Senior Technical Producer.

Competitive salary, benefits & build big things.

Your Job: You’ll work directly with our customers, their teams, and our engineers to distill requirements & develop project plans, design solutions, and keep projects running smoothly. You need to understand application development & project management and be able to communicate technology in human terms. A background as an engineer or some other technical role is preferred.

About Us: Citrusbyte is a highly specialized custom software development consultancy. We are brought in to build large software / technology projects such as ad networks, health platforms, business platforms and more. Our work is typically from concept through launch and beyond. We practice agile development, use small teams ( 2-6 ), and focus on developing a team of A players.

Apply: jobs@citrusbyte.com

Two points:

1. It's not always about the salary for employees. Some people just don't have the DNA to work at the type of company who gives the 120k salary ( assuming large-co ). They like small companies, being empowered, and all of the other benefits of being part of a team building something. They expect to get 70k and not recoup the difference because they consciously "buy" the lifestyle with the difference in pay. They are driven by the passion of the founder and want to be involved in building "something big". They can't miss the chance to get this experience that would take years in a big company.

This type of employee sees equity in two parts: First as an emotional connection that they're working on something they "own" and second as the dream of a potential payout that gives them the vision / hope of a great future. Both are important emotional motivators that enable the team to gather round a vision and kick-ass nights and weekends to make something valuable.

A startup employee who says, "4%, so you need to sell at 15m or i'm out of here…" is the same person who says they should pay less for electricity because they didn't use as many lights as you. That employee shouldn't be choosing to work at startup A because they offered offered 5k more salary and 1% more equity than startup B. Which one are they passionate about? Where do they want to spend their life for the next few years? The equity is just gambling.

2. Not all founders will share the risk equally even if they all work full-time, quit their jobs, etc.

Scenario A: Founder A is a serial entrepreneur with a few M in the bank while Founder B is an engineer who needs a paycheck. Does Founder A give B 50% or does he just pay him for his work as employee #1?

Scenario B: Founder A doesn't take salary from the company but has a consulting gig that pays 10k/month and takes ~4/hrs a week. Founder B is wealthy and doesn't take salary but doesn't need to spend 4-8 hours a week doing "other obligations" and says Founder A isn't dedicated.

Things get more complex once people become serial entrepreneurs.

The reasoning and analysis is so silly I would question all of the thinking that went into this article:

“The real problem with PRT in dense urban areas is parking. During non-rush hour periods, what do you do with 333k vehicles?”

What a huge face-palm.

This refers to his statement about NY where he naively assumed the most sensible analysis was one that replaced the existing subway completely with PRT. - Then, magically, every person would go buy a car ( why? ) - All those people would need parking spots at the PRT stations that are within minutes walking distance - He designed a one-size-fits-all solution instead of recognizing that a combination of several specialized solutions is even possible.

You're not going to replace the ultimate freedom cars give you, nor the huge capacity the subway provides. PRT does something else - it provides a supplemental point-to-point transportation.

Also, the peak demand measurement is flawed because it is based on existing infrastructure. X cars go on a freeway / hour. Y people go on a subway / hour. But you are forced to use the 10 freeway because its the only infrastructure that provides for MANY routes [ santa monica - downtown ], [ culver city - hollywood ] and so on. Considering the "mesh" feature of PRT requires an analysis instead of starting and ending destinations - not the infrastructure those trips used!

was thinking "redis" the whole time.