abi

5 ms·

1.
by abi 2mo ago
Fable 5 is included in the Claude Max subscription. I've gotten close to the limits this week and last but haven't hit it yet.
2.
by abi 3mo ago
It's worse than that. It's 5-4 on the read of the constitution since Kavanaugh doesn't think it violates the 14th amendment, just a federal statute.
3.
by abi 3mo ago
I think until Fable, Claude's vision was significantly worse than GPT and Gemini in my personal experience. I eval almost every vision model since I work on screenshot to code conversion project: https://github.com/abi&
4.
Show HN: Lilo – An open source personal AI assistant that lives in Telegram
2 points by abi 4mo ago|1 comments
5.
by abi 4mo ago
Your links are broken FYI.
6.
Show HN: 1-800-CODER, macOS app where you call an AI developer to edit your page
1 points by abi 4mo ago|0 comments
7.
by abi 4mo ago
https://github.com/abi/lilo I’m working on Lilo, a Telegram AI agent that can remember things, store files, track your TODOs, manage your calendar, conduct research, build apps, send you reminders and monitor things fo
8.
by abi 4mo ago
What benefit would it truly provide? Companies would simply say they need to cut costs to maximize shareholder value, which is no different than what happened here.
9.
Your AI assistant doesn't belong in your inbox(abiraja.com)
3 points by abi 5mo ago|0 comments
10.
by abi 5mo ago
No, we mostly spent our time on data structures and algorithms.
11.
by abi 5mo ago
Ugh, good point.
12.
Show HN: Lilo – a self-hosted, open-source intelligent personal OS(github.com)
7 points by abi 5mo ago|4 comments
13.
by abi 5mo ago
Usually, those get released a few weeks later.
14.
by abi 5mo ago
I'm quite confused by this article. If you persist conversation history in a database, and have all agentic turns run on the server, and merely listen to the streaming events/history via a websocket on the client, this is easily a
15.
by abi 5mo ago
Codex is great.
16.
by abi 2y ago
Exactly.
17.
by abi 2y ago
I’ve experienced that issue as well. Clearing the cache and redownloading seemed to fix it for me. It’s an issue with the upstream library tvmjs that I need to dig deeper into. You should be totally fine on a 32gb system.
18.
by abi 2y ago
Use secret llama in a incognito window. Turn off the Internet and close the window when done.
19.
by abi 2y ago
Thanks for the bug report. Yeah, it’s a bug with not resetting the state properly when new chat is clicked. Will fix tomorrow. Chat history shouldn’t be hard to add with local storage and Indexed DB.
20.
by abi 2y ago
Yes. Web-llm is a wrapper of tvmjs: https://github.com/apache/tvm Just wrappers all the way down
21.
by abi 2y ago
Appreciate the kind words :)
22.
by abi 2y ago
Window AI ( https://windowai.io/ ) is an attempt to do something like this with a browser extension.
23.
by abi 2y ago
Try to switch models to something other than tinyllama (default only because it’s the fastest to load). Mistral and Llama 3 are great.
24.
by abi 2y ago
This is amazing. Thanks both for sharing your stories. Made my day.
25.
by abi 2y ago
Well, it should be possible to just drag and drop a file/folder
26.
by abi 2y ago
Just bought the domain a couple of hours ago so DNS might not have propagated. Try back tomorrow or download and install it from GitHub (it’s just 2 steps)
27.
by abi 2y ago
Would love to. It uses MLC AIs webllm so just need to convert it to that format.
28.
by abi 2y ago
Yes, it only starts the download after you send the first message so visiting the site won’t use up any space. Approx sizes are listed in the GitHub README. Models are stored in indexeddb and will be managed by the browser. Might get evicte
29.
Show HN: I built a free in-browser Llama 3 chatbot powered by WebGPU(github.com)
547 points by abi 2y ago|139 comments
30.
by abi 2y ago
Is the model being released?

More ›