OpenAi Guides and resources

• GPT Best Practices By Open Ai.
• OpenAi Cookbook on Github
• GPT Function Calling
• Use playgrounds to test stuff
• Function Call Explained

Software industry and AI

• (AlphaDev discovers faster sorting algorithms)[https://www.deepmind.com/blog/alphadev-discovers-faster-sorting-algorithms?utm_source=noise.beehiiv.com&utm_medium=newsletter&utm_campaign=sunday-signal-25] “AlphaDev, an artificial intelligence (AI) system that uses reinforcement learning to discover enhanced computer science algorithms – surpassing those honed by scientists and engineers over decades.”
• The Evolution of Digital Interactions: From UX to MX By David Poblador, on Keep it boring. " AI becomes our primary point of interaction for services we currently access through apps or websites. It’s the next natural evolution of technology – a future that brings us not only convenience but also a level of personalized service that apps and websites alone cannot provide."
• megaguide of ai tools (most of them api calls to GPT

Useful or interesting Models

Huggingface’s Llama implementation

The OpenLLaMA repository is an open-source reproduction of Meta AI’s LLaMA large language model. It includes 3B, 7B, and 13B models trained on 1T tokens, with PyTorch and JAX weights of pre-trained OpenLLaMA models available. The repository also provides evaluation results and comparisons with the original LLaMA models. The weights are released in two formats: EasyLM format for use with the EasyLM framework, and PyTorch format for use with the Hugging Face transformers library. Both the training framework and the checkpoint weights are licensed under the Apache 2.0 license. open llama 13b

Datasets

Magic Brush

MagicBrush is a large-scale, manually-annotated dataset for instruction-guided image editing. It includes diverse editing scenarios and consists of 10,000 source image, instruction, and target image triples. The InstructPix2Pix model, fine-tuned on MagicBrush, has shown improved performance. Extensive testing reveals the challenging nature of the dataset and the gap between current image editing baselines and real-world needs. link to magic brush repo