agunapal

4 ms·

1.
by agunapal 5mo ago
My first thought after reading the blog was, let me share the blog with Claude and ask it how bots can circumvent this. imo AI bots have significantly affected OSS and we need better qualitative measures to define success
2.
Most teams optimize the prompt. Agentic systems have more moving parts(aevyra.ai)
3 points by agunapal 5mo ago|0 comments
3.
Show HN: Verdict – model evals on your own data, not someone else's benchmark(github.com)
2 points by agunapal 5mo ago|0 comments
4.
by agunapal 5mo ago
Yes, one can easily setup agents to bump up the stars, increase pip downloads etc
5.
by agunapal 5mo ago
I think it comes down to "Is the juice worth the squeeze" As someone who worked for a large organization maintaining an OSS project, one issue I faced was how do you show impact? We used to have many organizations really love and
6.
by agunapal 5mo ago
Here is a paper from few years ago where they talk about 7x speed increase, which equates to savings. https://arxiv.org/abs/2101.03961
7.
by agunapal 5mo ago
Do you mean model sharding?
8.
by agunapal 5mo ago
Very competitive price for the speed and intelligence being offered!
9.
by agunapal 5mo ago
Nvidia had the first movers advantage. Nvidia spent so many years perfecting CUDA to work well with PyTorch. Before ROCM, there was only CUDA. There were so many developers building their use cases on top of PyTorch+CUDA, and bringing all t
10.
by agunapal 5mo ago
If you really think about why MoE came into existence, its to save significant cost during training, I don't think there was any concrete evidence of performance gains for comparable MoE vs dense models. Over the years, I believe all