Fernando Velasco Lozano
@fermaat.bsky.social
Father. Mathematician. Machine Learning person. Research, Deep Learning, Tolkien, Sanderson and Atleti
I've been playing with CrewAI to build agents and honestly? Pretty cool framework overall 👇
January 25, 2025 at 11:04 AM
I've been playing with CrewAI to build agents and honestly? Pretty cool framework overall 👇
Here we go again!
December 6, 2024 at 10:21 AM
Here we go again!
The idea of LoRA is so beautiful. Many of you would have probably used it or at least heard of it. This paper was published some days ago and it is a very interesting study on how it differs from regular pre-training: LORA VS FULL FINE-TUNING. Paper Link: https://buff.ly/3CPkebH
buff.ly
November 23, 2024 at 11:19 AM
The idea of LoRA is so beautiful. Many of you would have probably used it or at least heard of it. This paper was published some days ago and it is a very interesting study on how it differs from regular pre-training: LORA VS FULL FINE-TUNING. Paper Link: https://buff.ly/3CPkebH
A couple of weeks ago marked the anniversary of BERT being open-sourced - one of the most important moments in ML history. BERT’s moment was a true revolution: it excelled at question answering, NER, sentence classification... basically everything we could throw at it back then!
November 17, 2024 at 7:41 AM
A couple of weeks ago marked the anniversary of BERT being open-sourced - one of the most important moments in ML history. BERT’s moment was a true revolution: it excelled at question answering, NER, sentence classification... basically everything we could throw at it back then!