Can you fine tune tesseract on a local hand writing dataset ? Or insert it in context like a pre-prompt ?
Can you fine tune tesseract on a local hand writing dataset ? Or insert it in context like a pre-prompt ?
What’s the 411 on steamunlocked ? Just malware ? Can we just get the cracks like in the old days ?
Sounds great, I could use voidtool everything content: search
I would prefer not to save and tags tabs 500 times per day. It’s easier to let them accumulate and handle them all in memory.
500 tab save and tag per day is too much labour, I would spend half my day just fiddling and sorting bookmarks !
Hint: stores are flammable
If that happens, I will force a 1000x refund, per instance.
Easy legislative fix. If the price is changed after the customer picked it up but before they checkout, store must refund item 1000x.
I’ve researching that and it seems the bottleneck is going to be transfering the tab inner information to secondary storage software. This is often a multi step process and also imperfect. With many website expressly frustrating this attempt by deleting and reloading data which is out of sight.
For instance trying to archive a facebook thread. As you scroll down the thread, it loads tge text ahead, but it also delete a few pages behind.
I’m not sure tab data can be expected to translate reliably to another store systen. It might have to stay in the browser.
Best I could figure so far is a rolling video screenshot, but that makes the data huge and difficult and imprecise to search as you now have to OCR evety frame to make it searchable again.
Cool I would love to navigate my data in a manner similar to this. However not obsidian, I am in the process of de-googling and I have severe cloud fatigue. But maybe QOwnNotes
I’m hoping something like Archivebox or squid or some other software can help me, autodump everything in a way that will become accessible to these second party data management software. Hopefully in a manner as transparent as opening a tab.
No, I have to setup all the tabs in just the right way. Then for each tabs it gets the price and shipping information I paste that into excel Combine the total together and sort with ascending price Then I repeat that for every quantity value for 1,2,3,4,5,7,10,15,20,25,50,75,100 Then I find the minimum quantity to get the best price.
This is because if you go to the website and just ask “order by price” it either hides most results, or straight up lies and still place them out of order. It also lies about the shipping cost. But it can’t lie on the last page before clicking buy.
I expect the internet to continue becoming more deceptive and manipulative in this manner, my method is almost not good enough. If my tools don’t continue to evolve it will simply become impossible to find the best price for anything. It will all become an endless maze where they measure how much mental stamina you’re willing to waste to save another dollar. At that point the price of things will become whatever the maximum you individually will bear.
You dream to small Bookmarks suck and are cumbersome They sucked in 1996 and they still suck today ! Bookmarks have apparently been a crutch to make the browser more usable. Like for instance, instead of discarding a whole tab, keep a text index of the html body and make that searchable. But no, it’s an all of nothing thing, either 2gb of youtube javascript per tab, or we only keep URL and tab title.
Also, you don’t actually need to bring a solution to the table just to say “this thing is not working right” You don’t have to be a mechanic to say “the car is broken” You don’t have to be a doctor to say “this person is sick”
Clearly my message just need to be said over and over until it gets implemented. It is obvious where browsers are going. A total web awareness platform that remembers everything you’ve ever seen. There will be infinite tabs and a local llm will know it all 7 ways from sunday “Firefox, write a song about the 500 first tabs I’ve seen in June 2017, in the style of a broadway musical”
I needed to buy 15x CH9121 and that was the difference between 15$ each to 4.5$ each.
Yes, I find that it identical to closing a tab. I never go in the bookmarks manager after. It is very clunky to use, it adds extra steps compared to keeping the tab open. At that point, it’s usually easier to use google to find it again, since at least google can search text inside the page, not just the title. I do occasionally dump my thousands of tabs into the bookmarks managers, in a single unusable folder. It hasn’t yet happenned that one of these tabs was retreived. But I hope in the future that I could dump all these tabs into another piece of software that will fetch all the tab’s body data and allow me to search it all with a local LLM based search like “using my bookmarks, create one browser window with all URLs on the topic of the 7 megahertz maser” We’re close but not there yet.
Zotero
I like the sound of that, thanks !
browser hygiene habits
You used that term, and frankly I recoil a bit a this term because of the implication that it’s not a deficiency of the software but that it’s the users who are wrong.
Still, I typed in the phrase into chatgpt
And I see “reading lists” as an alternative to bookmarks (that I find to be, straight up unusable)
So I found this reading list addon give a try.
https://addons.mozilla.org/en-GB/firefox/addon/reading_list/
I have a very specific use for a “reading list”, which I take to be something like a FIFO stack of links. And that would be going through youtube videos.
Putting this in case someone else is reading this thread looking for answers.
However, it’s a side bar thing, and you have to add links one at a time, can’t select multiple tabs and add them
As for opening 500+ tabs to buy a thing.
You do know that sellers now use algorithmic pricing and often there will be hundreds of sellers for the same thing.
Plus the price will be obfuscated with various artifices that all have to be overcome to find the best seller with the best price.
Defeating all of that means openning a shit-ton of tabs.
Here’s an example of the process I’ve designed for aliexpress
https://github.com/igorlogius/gather-from-tabs/discussions/8
I mean, look at how much data a youtube tab actually download, versus how much it occupies in memory. I think the strict memory isolation between tabs, so that one tab crash doesn’t take down the entire browser, has become uneconomical. I think combining some tab memory. Especially tabs of the same websites, especially their libraries, would greatly reduce the memory consumption and probably overall speed. I rarely ever get crashes until I bust both my ram and swap. I would sacrifice some tab isolation to get some memory back.
Yes the ;) variety
Yes, files that go back to 1996 when I first got online. Much older stuff that I got afterwards
I was using Kodi and I am switching to Emby.
Various renamers
https://picard-docs.musicbrainz.org/en/config/options_filerenaming.html
https://github.com/mobeigi/filebot
and many custom bash and batch scripts
Yes, notepad has “search in all open files” which would be great in firefox, “search in all tabs” and then it shows you a tab list with the search text in context with excerpts, kind of like how google does. Then with one click you could jump to that place in the text in that tab.
I found the following It migth be possible and affordable
https://konfuzio.com/en/tesseract/
https://github.com/Matleo/Tesseract_fine_tuning_training
https://groups.google.com/g/tesseract-ocr/c/ZLOZpW1fD6I/m/B1Ponc0VBAAJ
https://arcruz0.github.io/posts/finetuning-tess/