Viewing profile — ibuildthings

Talk on optimizing matrix multiplication with Triton kernels, focusing on low-bit processing and efficient quantization for high-performance AI models.

Aana SDK is an open-source toolkit for building cutting-edge multimodal AI applications: https://github.com/mobiusml/aana_sdk It addresses key challenges in multimodal AI developme…

Currently leading in LLM benchmarks among the 3B categories of models

We are releasing new 2-bit Mixtral models. These ones use a mixed HQQ 4-bit/2-bit configuration, resulting in a significantly improved model (ppl 4.69 vs. 5.90) with a negligible 0…

We are releasing 2-bit and 4-bit quantized versions of Mixtral utilizing the HQQ method that we just published https://mobiusml.github.io/hqq_blog/ and https://github.com/mobiusml/…

Sharing our work on model quantization. - Blog: https://mobiusml.github.io/hqq_blog/ - Code: https://github.com/mobiusml/hqq - Models: https://huggingface.co/mobiuslabsgmbh/ No dat…

Github repo should be visible now. It is not distilling the model, it is reducing the model weights on the fly and uses LoRA for training/fine-tuning. After the training phase, we …

I'm sharing a blog post https://mobiusml.github.io/low-rank-llama2/ on our approach to pruning the Llama2 model by leveraging low-rank structures. In a nutshell, we've managed to r…

In this particular case, why is testing/demonstrating on diverse set of individuals a bad thing ? A personal anecdote is that a few years back the automatic door sensors in my univ…

While agreeing to the general principle, incentive structures are wired quiet differently in academia vs end consumer oriented gig/service industries. Publications ( number, when, …

First principles of doing a PhD and taking up an industrial jobs are quite different, which this article sidesteps. I am talking from the perspective of someone who did a PhD, post…

Author here. Just to clarify, motive of this work is to ease curation, with the massive amount of content being created; but by no means an attempt at creativity or originality. Th…

The author here. I used the term "understanding", not as in machines understanding the images, but more as scientific attempt in understanding aesthetics. ( "empowering me to devel…

I agree with you completely about "bad industry-focused research". It serves no end. My question is that, is it just a reflection of mediocracy and gaming/dishonesty being everywhe…

Pushing away from color lines is very easy. For negative values of lambda in Eq.1 of the paper, (i.e. reverse the cost for color lines ), the optimization tries to push the puzzle …

One of the authors here. Since it is a optimization, the difficulty can be controlled as a parameter ( the lambda parameter in Eq.1 in the paper ). But you are right, some of the p…