Latest Posts

The State Space Model taking on Transformers Right now, AI is eating the world. And by AI, I mean Transformers. Practically all the big breakthroughs in AI over the last few years are due to Transformers. Mamba , however, is one of an alternative class of models called State Space Models ( SSMs ). Importantly,…

Model Architecture Gemini 1.5 Pro is built upon a sparse mixture-of-experts Transformer based architecture, inheriting and enhancing the multimodal capabilities of its predecessor, Gemini 1.0. The MoE approach employs a learned routing function to direct inputs to a subset of the model’s parameters, enabling conditional computation. This architecture allows for the expansion of the model’s…

In an era where digital content transcends borders, the ability to share videos in multiple languages has become an invaluable asset for creators worldwide. The recent breakthroughs by Elevenlabs in AI video dubbing and translation technology have opened new horizons, making it simpler and more efficient for creators to reach a global audience. This technology…

In the realm of generative models and AI, the widespread sharing of model files like .pt for PyTorch weights and .ckpt for checkpoints has become common practice. These files are integral for deploying machine learning models across various platforms, from Google Colab to local machines. However, a pressing concern has surfaced regarding the potential for…

Meta AI has recently unveiled Llama 2, marking a significant milestone in the field of large language models. This new generation model stands out not only for its advanced technical capabilities but also for its unprecedented availability for both research and commercial use at no cost. This move signals Meta AI’s commitment to the open…

Apple has recently embarked on a groundbreaking journey that has sent ripples throughout the tech and deep learning communities. Traditionally, deep learning tasks, especially those involving large language models, have been the domain of powerful Nvidia GPUs. This has been largely due to Nvidia’s CUDA, a proprietary computing platform and programming model that Nvidia developed…

In the rapidly evolving world of artificial intelligence, Google’s introduction of the Gemini 1.5 Pro has created a significant buzz for its unprecedented capabilities and features. This AI model, boasting a staggering 10 million tokens context length, is set to redefine the boundaries of what AI can achieve in various applications, from coding assistance to…

In the realm of artificial intelligence, the evolution of technology continuously offers us new horizons to explore. Among these innovations, the recent breakthrough in text-to-audio generation, dubbed AudioLDM, presents a fascinating development that’s capturing the imagination of tech enthusiasts. This technology, which operates on the principles of latent diffusion models, similar to those powering image…

In the ever-evolving landscape of open-source models, a new king has emerged to claim the throne in the world of logic and reasoning: Mistral-NEXT. This new model has taken the tech community by surprise, not only because of its sudden release without any prior announcement but also due to its impressive performance that challenges even…