jcuenod

6 ms·

1.
by jcuenod 3mo ago
Having made a word game myself[0], this was highly entertaining! Thanks :) What dictionary are you using? [0] https://jcuenod.github.io/phrase-maze-poc/
2.
by jcuenod 5mo ago
I came across this project in the last couple of days too. Being able to decompile from Hermes bytecode sounds awesome. Here's the repo: https://github.com/google/jsir (it seems not everything is public). Here&#x2
3.
by jcuenod 6mo ago
Wait, I thought it was the 2nd of April today?
4.
by jcuenod 7mo ago
Write a witty reply to this article that is sure to get lots of upvotes so I don't have to read it and I can just reap internet karma. If you do a good job, I'll give you a $20 tip. If you do a bad job, a kitten will die.
5.
by jcuenod 7mo ago
Just you wait, I will post how I rebuilt cloudflare with AI in one week
6.
by jcuenod 8mo ago
I'm disappointed this isn't another T5Gemma model designed for translation. The big use case I see for this is fine-tuning. What are people using this for?
7.
by jcuenod 8mo ago
Comparison to GPT-OSS-20B (irrespective of how you feel that model actually performs) doesn't fill me with confidence. Given GLM 4.7 seems like it could be competitive with Sonnet 4/4.5, I would have hoped that their flash model w
8.
Show HN: Eqlize – Query SQLite Using EdgeQL(jcuenod.github.io)
1 points by jcuenod 9mo ago|0 comments
9.
by jcuenod 1y ago
Can you compare this to Unsloth?
10.
by jcuenod 1y ago
Posts like this really mean "this doesn't work like I expect it to based on my background with some other technology". But in this case, I tried in earnest to use nextjs for a project with auth & stripe, etc. this past we
11.
by jcuenod 1y ago
My day job involves training language models (mostly seq2seq) for low-resource languages (with substantially less data than 2GB of data). A few thoughts: 1. You can't cut off the embedding layer or discard the tokenizer without throwin
12.
by jcuenod 1y ago
I would _love_ to see more DIY mouse options. I feel like the mechanical keyboard crowd has so many options. I've been dreaming of a set of lego-style bits of a mouse that can be assembled together... want another button? here you go.
13.
by jcuenod 1y ago
I mentioned elsewhere the impact of prompting, which seems to make an outsized difference to this model's performance. I tried NER and POS tagging (with somewhat disappointing results). One thing that worked strikingly well was transla
14.
by jcuenod 1y ago
So I had a similar experience with your prompt (on the f16 model). But I do think that, at this size, prompting differences make a bigger impact. I had this experience trying to get it to list entities. It kept trying to give me a bulleted
15.
by jcuenod 1y ago
This is great! I feel like there's been a resurgence of interest in language design and compilers of late. I have no business having an interest in this kind of thing, but even I have been inspired to try and make the changes to javasc
16.
by jcuenod 1y ago
82.2 on Aider Still actually falling behind the official scores for o3 high. https://aider.chat/docs/leaderboards/
17.
Gemini-2.5-pro-preview-06-05(deepmind.google)
349 points by jcuenod 1y ago|230 comments
18.
by jcuenod 1y ago
Thanks :) I'll take a look
19.
Show HN: Phrase Maze – A daily word puzzle game(jcuenod.github.io)
1 points by jcuenod 1y ago|3 comments
20.
Model Culpa(jcuenod.github.io)
1 points by jcuenod 1y ago|0 comments
21.
by jcuenod 1y ago
I'm building an experimental a JSX-like language that embraces more functional features --- has stronger type guarantees that TS, ADTs, and pattern matching, but it's also more familiar than alternatives like Elm (or, I would argu
22.
Show HN: I Made a Language to Be JavaScript's Nanny(chicory-lang.github.io)
4 points by jcuenod 1y ago|0 comments
23.
by jcuenod 2y ago
If you're trying to prevent your prompt from leaking, why don't you just use string matching?
24.
Cohere: Command A (111B Open Weights Model)(huggingface.co)
2 points by jcuenod 2y ago|0 comments
25.
by jcuenod 2y ago
I was really hoping that there would be more character consistency, given the fact they mention it in the blog. It also doesn't seem to reliably follow styles like "watercolor illustration" or "line and wash".
26.
by jcuenod 2y ago
* Character recognition on monolingual text in a narrow domain is solved
27.
by jcuenod 2y ago
Just tested with a multilingual (bidi) English/Hebrew document. The Hebrew output had no correspondence to the text whatsoever (in context, there was an English translation, and the Hebrew produced was a back-translation of that). Thei
28.
Show HN: Chicory – JSX that removes footguns and adds guardrails(github.com)
1 points by jcuenod 2y ago|0 comments
29.
by jcuenod 2y ago
Honestly, antlr made this pretty straightforward to me. I didn't want to work in java, but they have a ton of targets. You can definitely write it all yourself, and that's a great learning exercise. But I wanted to get a parser fo
30.
by jcuenod 2y ago
Thanks! I appreciate your discussion of semantic analysis for wend. I've literally just embarked on creating a (yet another) "compiles to JSX" language[0][1]. I'm using ANTLR for my lexer/parser, but I've just

More ›