#models

This open-source app supports secure interaction with privately hosted AI models such as Llama 2 and Mistral on macOS, iOS, and visionOS. Compatible with Ollama for a seamless and private conversation experience, the app features text-to-speech, markdown support, and locally stored conversation histories. Voice prompts and image attachments enable a customizable interface. Enchanted's offline capabilities ensure privacy and efficiency across iOS devices, including Vision Pro.

TorchMultimodal is a PyTorch library designed for comprehensive multimodal multi-task model training. It provides modular fusion layers, adaptable datasets, and pretrained model classes while enabling integration with elements from the PyTorch framework. The library includes numerous examples for training, fine-tuning, and evaluating models on various multimodal tasks. Models such as ALBEF, BLIP-2, CLIP, and DALL-E 2 facilitate the replication of state-of-the-art research, providing a valuable resource for researchers and developers aiming to advance in multimodal model training.