dcastm

10 ms·

91.
by dcastm 3y ago
Tbh I’ve seen AutoGPT mentioned a lot, but haven’t actually seen anyone using it regularly
92.
by dcastm 3y ago
Not my case but I thought this might help. I wrote some tips on how to use LinkedIn effectively: https://dylancastillo.co/tips-for-standing-out-on-linkedin/ It’s based on published papers about its algorithm + personal
93.
Create your own code interpreter chatbot with Pyodide(dylancastillo.co)
1 points by dcastm 3y ago|0 comments
94.
Show HN: Make graphs and analyze your dataset in plain English(deepsheet.dylancastillo.co)
2 points by dcastm 3y ago|0 comments
95.
Semantic search with Elasticsearch and LangChain(dylancastillo.co)
1 points by dcastm 3y ago|0 comments
96.
by dcastm 3y ago
It runs three fairly slow queries on GPT-4, checks if the domains are available, that there aren't any registered trademark issues. It takes roughly 6 mins to run everything. Maybe a better choice would have been to show the results as
97.
Show HN: Generate memorable startup names using AI(namemancer.dylancastillo.co)
1 points by dcastm 3y ago|2 comments
98.
Show HN: Use ChatGPT on your desktop using an app built by ChatGPT(github.com)
1 points by dcastm 4y ago|1 comments
99.
by dcastm 4y ago
Oh nice! I'm curious to know how you did the finetuning. I've been thinking about it as well. You're correct about the approach. Sometimes it mixes the sources for whatever reason, but gotta investigate it more. For the sourc
100.
by dcastm 4y ago
Yes, it uses embeddings (but not from OpenAI) as those are too expensive.
101.
by dcastm 4y ago
Sorry, wasn't really prepared for that much traffic and got hugged to death for a while :/
102.
Show HN: Get advice from a GPT3-based stoic philosopher(seneca.dylancastillo.co)
109 points by dcastm 4y ago|125 comments
103.
Show HN: Ask Seneca a question and get an answer with references to his writings(seneca.dylancastillo.co)
1 points by dcastm 4y ago|0 comments
104.
by dcastm 4y ago
Nice! free-genie is my own API. It also uses gpt, but my goal is to collect data and then finetune a model with it.
105.
by dcastm 4y ago
What's wrong about it? It seems to get the result I want, but not sure if I'm missing something.
106.
by dcastm 4y ago
Haven’t used it. Will check it out, thanks!
107.
by dcastm 4y ago
Great idea. Providing links to the relevant manpages makes a lot or sense.
108.
by dcastm 4y ago
It won’t run any command unless you explicitly tell it to do so. But yes I’d also like to provide some sort of warning if a certain command “looks dangerous”. I’ve been thinking about what's the best approach.
109.
by dcastm 4y ago
Thanks! I thought of using a shorter name but the ones I liked were already used in pypi. You can always add `alias genie=shell-genie` to your .bashrc or .zshrc file :)
110.
Show HN: Interact with the terminal in plain English using GPT-3(github.com)
90 points by dcastm 4y ago|27 comments
111.
by dcastm 4y ago
Cool idea but seems like the kind of product bound to have the same fate as Google glasses or Segway.
112.
Show HN: An interactive data viz cookbook built with PyScript and Pyodide(dataviz.dylancastillo.co)
1 points by dcastm 4y ago|0 comments
113.
by dcastm 4y ago
Note that the guy who submitted the images to the contest created hundreds of images, and spent weeks finetuning the prompts and curating the results. Midjourney/Stable Diffusion/DALLE are doing to Photoshop what it did to traditi
114.
Show HN: I made an interactive pandas cheat sheet using PyScript(pandas.dylancastillo.co)
1 points by dcastm 4y ago|0 comments
115.
by dcastm 4y ago
In the M5 competition[1], most winning solutions used LightGBM. So ML beat classical. Just a couple of the winning solutions used DL. [1] https://www.sciencedirect.com/science/article/pii/S016920702...
116.
by dcastm 4y ago
Looks like she goat herself in trouble
117.
by dcastm 4y ago
Thank you!
118.
by dcastm 4y ago
Thank you for all the ideas!
119.
by dcastm 4y ago
I understand what you mean. Gonna think this more thoroughly. Thanks for the feedback.
120.
by dcastm 4y ago
You're right, I should make it easier for users to share the game. I have a very small audience in twitter, and don't tweet that often, so sharing it that way wasn't really making a difference.

More ›