mufasachan

6 ms·

1.
by mufasachan 4mo ago
Insightful comment ahead: Is the gravity set very high or am I getting too old to play Flappy Bird with Transformers?
2.
by mufasachan 6mo ago
The blog is technical, technical terms in the TL;DR seems relevant to me.
3.
by mufasachan 6mo ago
Me too, I set up a WG tunnel to access this.
4.
by mufasachan 1y ago
Same experience, my workflow is to run the container from a podman run command, check it runs correctly, podlet to create a base container file, edit the container file (notably with volume and networks in other quadet file) and done (theor
5.
by mufasachan 1y ago
I appreciate the motivation of the author but the examples of this blog do not convince me at all. For every situation, there would be an alternative (with init or not) that addresses the design issues. The first example is solved with name
6.
by mufasachan 2y ago
My bad! As you said, this part of Python code bases is chaotic (:
7.
by mufasachan 2y ago
Eh, ML/scientific Python is large and not homogeneous. For code that should work on cluster, I would lean towards a Docker/container solution. For simpler dependancy use cases, pyenv/venv duo is alright. For some specific lib
8.
by mufasachan 2y ago
I do not know the people's background of a lot of comments here. They might have much more experiences than me with tensors. But, in my deep learning code and works, when I need to design an operation that involves a mix as little as 3
9.
by mufasachan 2y ago
Thank you for your answer. Some code base have chaotic execution path through a monolithic code base (by design). So, for these code bases, I heavily rely on the debugger. But I like your suggestion to use both of them, I think it's a
10.
by mufasachan 2y ago
Thanks for sharing your experience!
11.
by mufasachan 2y ago
I would be curious to know how you managed to do this. I really tried to do this but the tons of dev tools I am using was too much for transitioning to neovim for my daily work. Namely, I need a DAP, multiple dev tools (lsps, linters, forma
12.
by mufasachan 2y ago
fair comparison: tiananmen square versus yt-dl. Hilarious. (I 100% agree with you, just your comment gave me a funny smile. The comparison is relevant.)
13.
by mufasachan 2y ago
Thanks, my bad.
14.
Nvidia Project Digits – Nvidia Workstation(nvidia.com)
3 points by mufasachan 2y ago|2 comments
15.
by mufasachan 2y ago
Thanks for the details!
16.
by mufasachan 2y ago
Great benchmark, very interesting. Although, I am not sure about the extrapolation of the H200 from the lambda bench. From my understanding, Lambda and theirs bench used different models - LLama 405B and Mistral 123B - with different bench
17.
by mufasachan 2y ago
An inspiring project. I am looking forward to see some gloves connected to a VR device. I think that some cheap sensors, a bit of bayesian modelling and a calibration step can offer a proper realtime hand gesture tracking.* I am already pic
18.
by mufasachan 2y ago
Arguably, it's off-topic, though I agree with the point. Lebanon has been struck by poverty, and as a result, they might have far fewer choices when it comes to providers in general. Manufacturing within Lebanon or trading with neighbo
19.
by mufasachan 3y ago
> Although the author OCR’ed the SAT questions and believes that they weren’t in the training data I agree that the author of the tweet fairly underestimates the potential portion of OCR'ed contents in OpenAI's training data. I
20.
by mufasachan 3y ago
Disclaimer: This is just my intuition, I do not have knowledge about LoRA on small models. It's possible that does not work. LoRA (for Low Rank) benefits from the "small changes" introduced during finetuning of a model. The u
21.
by mufasachan 3y ago
One explaination https://x.com/yampeleg/status/1707127722743325106?s=46&t=Cxa... I would be curious what does he mean by "semi-automated system for detecting benchmark leaks. " though.
22.
by mufasachan 3y ago
A simple research on Scholar with "facial recognition liars" leads to a few niche articles about the topic. So it seems that facial recognition for liars is a thing. After a quick look on the articles you can see those methods are
23.
by mufasachan 3y ago
TL;DR: One is about distance in space, the other is about spread in space. KL-Divergence is not a metric, it's not symmetric. It's biased toward your reference distribution. Although, it gives an probabilistic / information v
24.
by mufasachan 3y ago
To be honest, OP seems to be at the beginning of its reading journey. His method is laborious. This "chain of summaries" seems to be hardly effective. I think the fractal method is close to be atomic - so a good one IMO. I would s
25.
Use deep learning models from other containers(stackoverflow.com)
1 points by mufasachan 3y ago|1 comments
26.
by mufasachan 3y ago
Everything is in the SO post. Also, this is my first SO question, if you have any remarks regarding it - putting aside the problem in itself - I am all ears.