hyuuu

at the sake of being obvious, do you have a tiny llm gating this decision and classifying and directing the task to its appropriate solution?

does the lite version have a faster token output? or time to first token?

one point the author seems to be missing as to why we need full self driving everywhere is in the logistics industry. We are talking about interstate driving, from point A to point B, in this case it might involve industria environments such as a truck stop, weighing, loading/unloading etc, something that requires decision making across different environments (urban, highway, dense residential etc). Tesla FSD has the real potential for this unlock, more than just a ride hailing use case of getting people from point A to point B.

i have been your user since the early days, I want to say congrats you guys, I have been recommending your framework to everyone. I appreciate the responsive support you gave me on discord (though in the end, my questions are already on the docs lol)

if you use node then you have the following for N+1, by having a tighter integration with the database, for ex: https://www.graphile.org/postgraphile/performance/#how-is-it...

or prisma

https://pothos-graphql.dev/docs/plugins/prisma/relations

it used to be the case that graphql requests are in POST, but not anymore: https://nearform.com/open-source/urql/docs/api/core/#:~:text... example for urql

on top of http level caching, you can do any type of caching (redis / fs / etc) just like a regular rest but at a granular level, for ex: user {comments(threadId: abc, page: 1, limit: 20) { body, postedAt} is requested and then cached, another request can come in thread(id: abc) {comments(page: 1, limit: 20) {body, postedAt} you can share the cache.

N+1 is solved by a tighter integration with the database, for ex: https://www.graphile.org/postgraphile/performance/#how-is-it... or maybe a more popular flavor using prisma: https://pothos-graphql.dev/docs/plugins/prisma/relations#:~:...

but of course, there is always the classic dataloader as well.

I am not saying that use graphql and all the problems will be solved, but i am saying that the problem that OP proposed has been solved in an arguably "better" way, as it does not tie the presentation (HTML) with the data for cases of multiplatform apps like web, or native apps.

N+1 is a solved problem at the framework level If GraphQL actually affects your performance, congratulations, your application is EXTREMELY popular, more so than Facebook, and they use graphql. There are also persisted queries etc.

Not sure about caching, if anything, graphql offers a more granular level of caching so it can be reused even more?

The only issue I see with graphql is the tooling makes it much harder to get it started on a new project, but the recent projects such as gql.tada makes it much easier, though still could be easier.

I was just going to say, all of this has been solved with graphql, elegantly.

link to the full pdf?

i have been looking for something like this, the closest I could find by googling was celery workflow, i think you should do better marketing, I didn't even realize that hatchet existed!

have you considered using remix?

I 100% agree with this, I think you can even extend this to any AI, in the end, IMO, as the llm is more commoditized, the surface of which the value is delivered will matter more

gemini flash is notorious for hallucinating the output of the OCR, be careful with it. For straight forward, semi-structured, low page count (under 5) it should perform well, but the more the context window is stretched the more the output gets more unreliable

It's a weird timing because I just launched https://dochq.io - ai document extraction where you can define what you need to get out your documents in plain English, I legitimately thought that this was going to be such a niche product but hell, there has been a very rapid rise for AI-based OCR lately, an article/tweet even went viral 2 weeks ago I think? About using Gemini to do OCR, fun times.

China / South Korea esport players will be the next super soldiers

sigh, for real..

could you elaborate on the multi modal aspect of this model?

how is it ridiculous? so far, reducing saturated fat intake has a direct correlation with lowering LDL that is a marker for cardio risk no?

i didnt realize that the diet im doing right now has a name, ornish! Just to clarify, so this diet actually works in preventing further damage?

do you use any agentic prompting techniques?

the view of the comments here seems to be quite negative for what meta is doing. Honest question, should they go to the route of openai and closed source + paid access instead? OpenAI or Claude seem to garner more positive views than llama open sourced.

where did you source your data?

does this integrate with django?

I have been using https://mantine.dev/ very polished and also has a very extensive collection of surprisingly useful hooks. Surprising because, whenever I try to create my own, 90% of the time mantine already has it that i can just import.

you should try driving with openpilot and compare it with FSD to see if they're equal since you are making this comparison

just make sure the clean energy does not come from China, US will tariff that to oblivion lol

this might be a stupid question, but how do the teams share the documents?

funny you mentioned that because this demographic is the same demographic that predicted the demise of FB (it has grown in orders of magnitudes), hating on Tesla (also has been growing), didn't believe the viability of Dropbox idea.

comparing it with a monitor is rather unfair, you have to bundle the computing along with it, not to mention the applications to make it an actually fair comparison. At that point, is it still ludacris?