shekhar101

8 ms·

1.
by shekhar101 6mo ago
https://github.com/jordanbaird/Ice Try this. I am not affiliated in anyway. It's my must have macos app.
2.
by shekhar101 6mo ago
Tried this and really liking it so far. Question - is there a diarization support in the tui app or any of the models MetalRt supports? Any plans to add it if not already supported?
3.
by shekhar101 11mo ago
Seems to be based on See through walls by MIT (2013)? Good job porting it to esp32. I was just looking a week ago to do the same thing - basically reproducing this work. I'll definitely try this. https://www.media.mit.edu&#x
4.
by shekhar101 1y ago
I was reading Xiaomi YU7 marketing page[0] yesterday and the NVIDIA AGX Thor stood out (says: NVIDIA DRIVE AGX Thor). I was wondering what it was and this showed up! Looks like it is (or a Drive variant of it) is already being used in newer
5.
by shekhar101 1y ago
Can someone (or OP) point me to a recipe to fine tune a model like this for natural language tasks like complicated NER or similar workflows? I tried finetuning Gemma3 270M when it came out last week without any success. A lot of tutorials
6.
by shekhar101 1y ago
FYI: owhisper pull whisper-cpp-large-turbo-q8 Failed to download model.ggml: Other error: Server does not support range requests. Got status: 200 OK But the bas
7.
by shekhar101 1y ago
I tried a picture with instructions and it says "something went wrong". I would love to try and see how well it works for my use case.
8.
Magnifier on Mac [video](youtube.com)
1 points by shekhar101 1y ago|0 comments
9.
by shekhar101 1y ago
Yes it is, however API keys from aistudio only allows pro-experimental model. So if I select gemini-pro, I will see this: "Gemini 2.5 Pro Preview doesn't have a free quota tier. Please use Gemini 2.5 Pro Experimental (models/
10.
by shekhar101 1y ago
Can you open up the options to use other model/versions, especially Gemini-2.5 pro experimental models available through aistudio? Would love to try this but gemini flash fails for even simple tasks. Example: I asked it to extract all
11.
Terence Tao on how we measure the cosmos [video](youtube.com)
1 points by shekhar101 1y ago|0 comments
12.
The Most Useful Thing AI Has Ever Done [video](youtube.com)
1 points by shekhar101 2y ago|1 comments
13.
by shekhar101 2y ago
This is really cool (and timely for me). Lovely work with the UX. No accounts, no nonsense. Kudos.
14.
by shekhar101 2y ago
> Spends so much time looking for UI design tools that they could've learned actual design in the meantime Ouch! This is really well done. :)
15.
by shekhar101 2y ago
I am RM2 user and this is useful for me. I signed up not expecting much but that digest designer is so well done! Kudos. Beautifully executed. Signed up and looking forward to the digests.
16.
by shekhar101 2y ago
India is around 94% electrified and seems to be highest amongst larger countries.
17.
by shekhar101 2y ago
Can't see pricing, can't even see demo video on the home page without logging in.
18.
by shekhar101 2y ago
Looks great! Do you mind talking about your tech stack? Do you build on top of Tessaract or do you use a custom model?
19.
Steroids Are Awesome [video](youtube.com)
3 points by shekhar101 2y ago|0 comments
20.
by shekhar101 2y ago
Tangential - I just want a decent (financial transaction) Table to text conversion that can retain the table structure well enough (e.g. merged cells) and have tried everything under the sun short of fine tuning my own model, including all
21.
by shekhar101 2y ago
Thank you very much!
22.
by shekhar101 2y ago
What's the name of the layout recorgniser model? I did not have a good experience extracting layout from tables, especially those without column boundaries (space instead of lines to demarcate boundaries)
23.
by shekhar101 2y ago
It's trivial to run a proxy server that routes all OpenAi calls to another LLM, even local ones. See litellm-proxy.
24.
Jeremy Howard: CUDA for Python Programmers [video](twitter.com)
63 points by shekhar101 3y ago|5 comments
25.
Coqui Is Shutting Down(coqui.ai)
1 points by shekhar101 3y ago|2 comments
26.
by shekhar101 3y ago
Explanation from Andrej karpathy makes sense on why: ''' "8x7B" name is a bit misleading because it is not all 7B params that are being 8x'd, only the FeedForward blocks in the Transformer are 8x'd, everyt
27.
Anthropic Claude for Google Sheets(docs.anthropic.com)
4 points by shekhar101 3y ago|1 comments
28.
by shekhar101 3y ago
Let me put a counter point to the opening remark. We do not (just) inherently trust people to do the right thing. Law and order, refined and (somewhat, however faultily) applied over the course of decades have programmed us to not do the wr
29.
by shekhar101 3y ago
Warning - The page played a loud music when I clicked on something. Most stuff seems to be under development - so what is it?
30.
Ask HN: What's the current best way to extract tables from PDFs?
4 points by shekhar101 3y ago|2 comments

More ›