text processing
-
Quill: Open Source Writing Assistant with Prompt Control
Read Full Article: Quill: Open Source Writing Assistant with Prompt Control
Quill is a streamlined open-source background writing assistant designed for users who want more control over prompt engineering. Inspired by Writing Tools, Quill removes certain features like screen capture and a separate chat window to focus on selected text processing, making it compatible with local language models. It allows users to configure parameters and inference settings, and supports any OpenAI-compatible API, such as Ollama and llama.cpp. The user interface is kept simple and readable, though some features from Writing Tools are omitted, which might be missed by some users. Currently, Quill is available only for Windows, and feedback is encouraged to improve its functionality. This matters as it provides writers with a customizable tool that enhances their writing process by integrating local language models and offering greater control over how prompts are managed.
-
Liquid AI’s LFM2.5: Compact Models for On-Device AI
Read Full Article: Liquid AI’s LFM2.5: Compact Models for On-Device AI
Liquid AI has unveiled LFM2.5, a compact AI model family designed for on-device and edge deployments, based on the LFM2 architecture. The family includes several variants like LFM2.5-1.2B-Base, LFM2.5-1.2B-Instruct, a Japanese optimized model, and vision and audio language models. These models are released as open weights on Hugging Face and are accessible via the LEAP platform. LFM2.5-1.2B-Instruct, the primary text model, demonstrates superior performance on benchmarks such as GPQA and MMLU Pro compared to other 1B class models, while the Japanese variant excels in localized tasks. The vision and audio models are optimized for real-world applications, improving over previous iterations in visual reasoning and audio processing tasks. This matters because it represents a significant advancement in deploying powerful AI models on devices with limited computational resources, enhancing accessibility and efficiency in real-world applications.