Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
samlhuillier
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
samlhuillier
1y ago
The output is literally a link, like typeform. And that link will render the design you've customized within the app
2.
▲
Banana Form: Typeform clone that let's you design forms with AI
(bananaform.app)
1 points
by
samlhuillier
1y ago
|
2 comments
3.
▲
by
samlhuillier
2y ago
I like to think of this like fine-tuning LLMs. When you fine-tune an LLM it doesn't pick up and memorise all of the training data. Rather it adjusts its weights/perspectives based on the content trained on/read.
4.
▲
France to fine patients €5 for missing GP appointments
(theguardian.com)
3 points
by
samlhuillier
2y ago
|
1 comments
5.
▲
by
samlhuillier
3y ago
Just markdown right now, plain text is coming
6.
▲
by
samlhuillier
3y ago
Not yet!
7.
▲
by
samlhuillier
3y ago
No catch pal
8.
▲
by
samlhuillier
3y ago
Thank you!
9.
▲
by
samlhuillier
3y ago
Interesting. Thanks for sharing will take a look!
10.
▲
by
samlhuillier
3y ago
Thanks for your feedback! - Multiple vaults is in fact in a PR right now: https://github.com/reorproject/reor/pull/28 - Manual linking is coming. - Minimizing the UI to chat is interesting. Right now I guess
11.
▲
by
samlhuillier
3y ago
Absolutely! Really respect the work you folks are doing.
12.
▲
by
samlhuillier
3y ago
Yes unfortunately not implemented yet. Will be coming soon though :)
13.
▲
by
samlhuillier
3y ago
Thank you for your feedback! Working hard on improving the chunking to improve related notes section. RAG is fairly naive right now, with lots of improvements coming in the next few weeks.
14.
▲
by
samlhuillier
3y ago
Two reasons: 1. The libraries I used to run models locally didn't work inside a plugin. 2. I believe AI is a fairly big paradigm shift that requires new software.
15.
▲
by
samlhuillier
3y ago
It's filesystem mapping 1:1. Basically the same thing Obsidian does when you open a vault. You can create new files with Reor, create directories and edit existing files. Chunking happens only in vector DB and everything is synced auto
16.
▲
by
samlhuillier
3y ago
I agree with this. In some cases, hard thinking and searching for things manually can really enhance understanding and build your knowledge. In other cases, particularly when ideating for example, you want to be given "inspiration"
17.
▲
by
samlhuillier
3y ago
If you can convert your One Note notes to markdown files then yes. On startup, you'll be asked to choose your vault directory - which needs to be a directory full of markdown files.
18.
▲
by
samlhuillier
3y ago
Absolutely! Would love your help.
19.
▲
Show HN: Reor – An AI note-taking app that runs models locally
(github.com)
411 points
by
samlhuillier
3y ago
|
102 comments
20.
▲
by
samlhuillier
3y ago
Times are changing. Human researchers will dedicate more and more time towards getting language models to work in desired ways rather than doing the research themselves. Language models will largely be the ones making "research" d
21.
▲
Top tech firms to let governments vet AI tools
(theguardian.com)
1 points
by
samlhuillier
3y ago
|
0 comments
22.
▲
by
samlhuillier
3y ago
ChatGPT hasn’t got the faintest idea about today’s news!
23.
▲
Show HN: AI News Analyst
(newsanalyst.vercel.app)
1 points
by
samlhuillier
3y ago
|
4 comments
24.
▲
Investigations into fine-tuning LLMs to be good at RAG
(ragntune.com)
2 points
by
samlhuillier
3y ago
|
0 comments
25.
▲
GPT 3.5 vs. Llama 2 fine-tuning: A Comprehensive Comparison
(ragntune.com)
47 points
by
samlhuillier
3y ago
|
12 comments
26.
▲
by
samlhuillier
3y ago
Legend
27.
▲
A simple guide to fine tuning Llama 2 on your own data
(brev.dev)
20 points
by
samlhuillier
3y ago
|
2 comments
28.
▲
by
samlhuillier
3y ago
Yes using the qlora adapter that hugging face provides with peft
29.
▲
by
samlhuillier
3y ago
Working on this now!
30.
▲
A simple guide to fine-tuning Llama 2
(brev.dev)
280 points
by
samlhuillier
3y ago
|
53 comments
More ›