I’m just in the beginning, but my plan is to use it to evaluate policy docs. There is so much context to keep up with, so any way to load more context into the analysis will be helpful. Learning how to add excel information in the analysis will also be a big step forward.

I will have to check out Mistral:) So far Qwen2.5 14B has been the best at providing analysis of my test scenario. But i guess an even higher parameter model will have its advantages.

Thank you! Very useful. I am, again, surprised how a better way of asking questions affects the answers almost as much as using a better model.

I need to look into flash attention! And if i understand you correctly a larger model of llama3.1 would be better prepared to handle a larger context window than a smaller llama3.1 model?

Thanks! I actually picked up the concept of context window, and from there how to create a modelfile, through one of the links provided earlier and it has made a huge difference. In your experience, would a small model like llama3.2 with a bigger context window be able to provide the same output as a big modem L, like qwen2.5:14b, with a more limited window? The bigger window obviously allow more data to be taken into account, but how does the model size compare?

Thank you for your detailed answer:) it’s 20 years and 2 kids since I last tried my hand at reading code, but I’m doing my best to catch up😊 Context window is a concept I picked up from your links which has provided me much help!

The problem I keep running into with that approach is that only the last page is actually summarised and some of the texts are… Longer.

Do you know of any nifty resources on how to create RAGs using ollama/webui? (Or even fine-tuning?). I’ve tried to set it up, but the documents provided doesn’t seem to be analysed properly.

I’m trying to get the LLM into reading/summarising a certain type of (wordy) files, and it seems the query prompt is limited to about 6k characters.

Well, that’s been the basis for some other products. AMD and Intel comes to mind😊 They both have IP the other need and historically Intel has been the dominant one, but now the tables have turned somewhat.

Had blocking news and access to information been in the cards, as you describe, there would be another discussion. This is not it. The closest this comes is to block a linkaggregator. One that has been deemed to violate the laws in its area of business and being reluctant to take steps to rectify the situation.

This being the supreme court doing it does bring up the question of democratic decision making, which famiously has been proven by other countries recently. Although they also gave their president the power to remove themselves from office, if I’ve understood that particular debacle correctly.

Well… Its built on statistics and statistical inference will return to the mean eventually. If all it ever gets to train on is closer and closer to the mean, there will be nothing left to work with. It will all be the average…

Well, there is a big discussion about nutrition among lupies, so maybe heuristics was on to something this time. On the other hand, in what group isn’t nutrition a topic?

An LLM once explained to me that it didn’t know, it simulated an answer. I found that descriptive.

Totally unlike the other fielded armies globally at the moment.

/S

No, we need to be able to keep two thoughts in our heads at the same time or we are bound to repeat the mistakes. Terror and oppression is terrible regardless of what the purpetrator and the victim are called.

A warning and belittlement. “Talk to your client”. The godfather wasn’t addressed.

Five years ago I was astonished that the U.S. navy was looking for renewable fuels thinking it would never pan out. Today I see alternatives which has nothing to do with fossils in the private sector. I see heavy transportation companies trying to find ways to electrify their fleet. I think those “grownups” will find themselves in a new reality despite them not wanting to.

And unfeasible? Three years ago maybe. Today we need to revise the time plans. I’m already working at implementing the goals set for 2030 and have started shifting focus towards the big elephant and the 2045 goals.

What? Won’t creditors accept real estate as security from a man found guilty of fraudulently overvalue his real estate in orde to use as security for the exact purpose of lending more money in a money lending scheme? Shocking!

I wonder, though: if the state of NY expropriate real estate to sell to highest bidder and those auctions fail to meet market value, won’t that create chain reactions? A) the creditor(s) which had that particular building as security will see the value of their asset turn to zero because of negligence from their customer “billionaire”, which should open for a case against Trump for both immediate repayment of the loan and subsequent damages, and B) send a clear signal to all other creditors that they need to secure their assets asap?

So, in a word… And I say this as a European non-english speaker, so I may have got meaning and culture all wrong… It would be unpatriotic?

Oh my… Last time I read pieces like this, the new architecture was called bulldozer.

It needs to be low, but positive and keept stable. If it’s to high it will be self sustaining and increasing, if it’s negative everything stalls. 2% seems to fit the bill.

There could be an argument that 4% would have been just as good, and had the rest of the world united on 4% it would*. However, it would not have changed anything in last year’s combat of inflation. The target would have been defended just as fiercely causing just as much collateral. Only the numbers would have been slightly different.

*Ignoring for a bit those countries that has had to fight to keep inflation up.