Latest posts 5
We have been curious about this for a long time! this resonates our findings - LoRA underperforms in settings that resemble pre-training, but matches FT in many cases.
We show that it is possible to reduce the serving cost with FT: arxiv.org/abs/2312.05215
We show that it is possible to reduce the serving cost with FT: arxiv.org/abs/2312.05215
LoRA makes fine-tuning more accessible, but it's unclear how it compares to full fine-tuning. We find that the performance often matches closely---more often than you might expect. In our latest Connectionism post, we share our experimental results and recommendations for LoRA.
thinkingmachines.ai/blog/lora/View quoted post on X
Finally! I had been helping build an internal serving system that aims to be cost-effective, heterogeneous and decentralized. Here’s my post reflecting on why: about.yao.sh/posts/swissai-…
(Could have done better, but next time :))
(Could have done better, but next time :))
about.yao.sh Reflection on Building Swiss AI Serving Reflection on Building Swiss AI Serving @EPFL , @ETH_en and #CSCS today released Apertus, Switzerland's first large-scale, multilingual language model (LLM). As a fully open LLM, it serves as a building block for developers and organizations to create their own applications: cscs.ch/science/comput… #Apertus #AIView quoted post on X
My reviews achieved 100% accuracy on the final results, I hope I did a decent job in helping review (I also particularly like the “reasons to accept” and “reasons to reject” setting!)
The list of accepted papers for COLM 2025 is now available here:View quoted post on X
colmweb.org/AcceptedPapers…
The papers will be made available around mid August, following the camera ready deadline
