Grigory SapunovGPT-4V is coming!The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)6 min read·Oct 5, 2023----
Grigory SapunovLLMs and a possible future for SearchThe recent surge of Generative AI has pushed the boundaries of what was previously thought possible.7 min read·Dec 22, 2022--1--1
Grigory SapunovOpenAI and the road to text-guided image generation: DALL·E, CLIP, GLIDE, DALL·E 2 (unCLIP)Let’s look at the evolution of text-guided image generation models from OpenAI, as there are more datapoints than just DALL·E and DALL·E 2.21 min read·May 1, 2022--1--1
Grigory SapunovFoundation ModelsIn August 2021 Stanford announced establishing the Center for Research on Foundation Models (CRFM) as part of the Stanford Institute for…13 min read·Nov 22, 2021----
Grigory SapunovHardware for Deep Learning. Part 4: ASICThis is a part about ASICs from the “Hardware for Deep Learning” series. The content of the series is here.49 min read·Jan 12, 2021--1--1
Grigory SapunovJAXJAX by Google Research is getting more and more popular. Deepmind recently announced they are using JAX to accelerate their research and…6 min read·Dec 20, 2020--1--1
Grigory SapunovGPT-3: Language Models are Few-Shot LearnersOpenAI just published a paper “Language Models are Few-Shot Learners” presenting a recent upgrade of their well-known GPT-2 model — the…11 min read·Jun 2, 2020----
Grigory SapunovFP64, FP32, FP16, BFLOAT16, TF32, and other members of the ZOOThere are many floating point formats you can hear about in the context of deep learning. Here is a summary of what are they about and…7 min read·May 16, 2020--2--2
Grigory SapunovAdaptive Computation Time (ACT) in Neural Networks [3/3]Part 3: ACT in Transformers9 min read·Jan 7, 2020----
Grigory SapunovAdaptive Computation Time (ACT) in Neural Networks [2/3]Part 2: ACT in Residual Networks7 min read·Jan 3, 2020----