Data science, career and other topics
20 July 2023
Meta's open-source LLM family (7B–70B parameters) with chat fine-tuning that matched or beat closed-source models on most benchmarks — a detailed look at the pretraining, RLHF, and safety techniques behind it.
17 July 2023
My review of the paper Scaling Autoregressive Multi-Modal Models Pretraining and Instruction Tuning
13 July 2023
My review of the paper UniverSeg Universal Medical Image Segmentation
10 July 2023
My review of the paper Recognize Anything A Strong Image Tagging Model
06 July 2023
My review of the paper Hiera A Hierarchical Vision Transformer without the Bells-and-Whistles
03 July 2023
My review of the paper Multilingual End to End Entity Linking
Type at least 2 characters to search...