pedrovhb

8 ms·

1.
by pedrovhb 13d ago
You'd use an incremental parser, they're specifically designed to be efficient for small edits. The big name is `tree-sitter`, but its output is compiled so on the web it requires some wasm shenanigans. For webpages, I'm a fa
2.
by pedrovhb 6mo ago
Very interesting, thanks for sharing. I'm curious about the focus on the language design and features with regards to agent orchestration vs. being a general purpose/ML architecture oriented language. The headline examples go &quo
3.
by pedrovhb 1y ago
Nice! Seems very useful if you can drop in and have everything work. Nitpicking a bit: it's not as much _rendering_ markdown as it's _syntax highlighting_ it. Another interesting approach there could be to use the CSS Custom Highl
4.
by pedrovhb 1y ago
You can presumably run code that calls `sys.settrace` for that. Which makes it somewhat underwhelming to realize that you pretty much could also do that before, but perhaps convenient that now you don't have to have the foresight to ha
5.
by pedrovhb 1y ago
As a Brazilian - Pix was a pleasant surprise, especially in that for once it feels like we're not lagging behind. It's convenient, free, instant transfers across banks. You can also easily create or programmatically generate QR co
6.
by pedrovhb 2y ago
> but most probably was training data to prevent deepseek from completing such prompts, evidenced by the `"finish_reason":"stop"` included in the span attributes As I understand, the finish reason being “stop” in API
7.
by pedrovhb 2y ago
I imagine it wouldn't necessarily require their opening of remote connections, just a misconfigured reverse proxy.
8.
by pedrovhb 2y ago
This seems centralized, though you can self-host it.
9.
by pedrovhb 2y ago
> [the extensive anti-reverse engineering measures are] more annoying than any financial app I've had, and I have 5 of them on my phone Ah, this reminds me of the Tuya app. I've done some ssl unpinning and mitm to see requests
10.
by pedrovhb 2y ago
That's very interesting! My first thought, looking at the webpage: "Huh, that's neat. I didn't know that painting software didn't even attempt to do color mixing beyond naive interpolation, though I guess it figures
11.
by pedrovhb 2y ago
Here's an idea: have the LLM output each comment with a "severity" score ranging from 0-100 or maybe a set of possible values ("trivial", "small", "high"). Let it get everything off of its chest
12.
by pedrovhb 2y ago
That's certainly possible, but it reminds me a bit of a similar thing I've seen in their UI that rhymes in a way that makes me think otherwise. In the code interpreter tool, you have a little preview of the "steps" it&#x
13.
by pedrovhb 2y ago
For what it's worth, as a primarily backend dev having ~recently started getting more deeply into frontend web, I have specifically noted in my head that the box model isn't too intuitive and in my inexperienced opinion, the defau
14.
by pedrovhb 2y ago
It does feel right to me, because it's not distilling the second model, and in fact the second model is not an image generation model at all, but a visual encoder. That is, it's a more "general purpose" model which speci
15.
by pedrovhb 2y ago
You may already be aware of it, but in case not - it sounds like tree-sitter-graph could be something you'd be interested in: https://docs.rs/tree-sitter-graph/latest/tree_sitter_graph/r... I haven'
16.
by pedrovhb 2y ago
Or by the definition that the ratio between consecutive fib numbers approaches Phi, just multiply by 1.618? Though at that point might as well just use the real conversion ratio. In other news, π² ≈ g.
17.
by pedrovhb 2y ago
+1 on feeling there's a lot of UX possibilities left on the table. Most seem to have accepted chat as the only means of using LLMs. In particular, I don't think most people realize that LLMs can be used in very powerful ways that
18.
by pedrovhb 2y ago
It does view RAW when compiled with the right flags. JXL too, interestingly. Managed to save a bunch of space on old photos (converting with cjxl, but which I wouldn't have done if I weren't able to see them somehow).
19.
by pedrovhb 2y ago
Here's an idea: recursively mount code files/projects. Use something like tree-sitter to extract class and function definitions and make each into a "file" within the directory representing the actual file. Need to get a
20.
by pedrovhb 2y ago
Have you tried asking it for a specific concrete length, like a number of words? I was also frustrated with concise answers when asking for long ones, but I found that the outputs improved significantly if I asked for e.g. 4000 words specif
21.
by pedrovhb 2y ago
My intuition is that a significant challenge for LLMs' ability to do arithmetics has to do with tokenization. For instance, `1654+73225` as per the OpenAI tokenizer tool breaks down into `165•4•+•732•25`, meaning the LLM is incapable o
22.
by pedrovhb 2y ago
Interesting! Did you try the function calling API? I feel you with the line number troubles, it's hard to get something consistent there. Using diffs with GPT-4 isn't much better in my experience; I didn't extensively test th
23.
by pedrovhb 2y ago
I thought of something similar these days but with a different approach - rather than settrace, it would use a subclass of bdb.Bdb (the standard library base debugger, on top of which Pdb is built) to actually have the LLM run a real debugg
24.
by pedrovhb 3y ago
How is suggesting the use of iterators and named tuples related to creating domain specific languages? If anything I'd say they're a much more generic and universally recognizable approach than having users subclass `AssistantEven
25.
by pedrovhb 3y ago
For all the brilliance in the AI and infra departments of OpenAI, their official Python library (which is the flagship one as I understand) feels pretty unidiomatic, designed without much thought for common patterns in the language. 2012 Ja
26.
by pedrovhb 3y ago
Meta seems to actually be taking all the right steps in how they're contributing to open source AI research. Is this a "commodotize your complement" kind of situation?
27.
by pedrovhb 3y ago
Technically a function body needs at least one statement. A docstring is just an expression statement (a string), so a function definition with just a docstring is synctatically valid Python. I've seen people say multiline string liter
28.
by pedrovhb 3y ago
Close. Temperature is the coefficient of a term in a formula that adjusts how likely the system is to pick a next token (word/subword) which it thinks isn't as likely to happen next as the top choice. When temperature is 0, the ef
29.
by pedrovhb 3y ago
Not sure I understand your point. LSP is a protocol and tree-sitter is a parser generator. They're kind of orthogonal concepts; a tree-sitter parser couldn't ever be used directly in place of an LSP server, but an LSP server may w
30.
by pedrovhb 3y ago
That's an interesting idea. Analogous to how LLMs are simply "text predictors" but end up having to learn a model of language and the world to correctly predict cohesive text, it makes sense that "video predictors"

More ›