github llama.cpp · 1h ago · inference

llama.cpp b10271 adds per-conversation working directory and Windows path support

from English

The llama.cpp b10271 release introduces a per-conversation working directory feature to the UI, allowing users to set and navigate directories independently for each chat session.

The server extends file_glob_search to support UI pickers with improved path navigation.
Path-like queries starting with / or ~ are treated as directory navigation rather than glob-matching.
Windows drive roots (e.g., C:) and UNC shares are now recognized for path navigation.
Backslashes in queries are rewritten to forward slashes only when a Windows root is present.
The working directory picker is decoupled into utils and sub-components for better maintainability.
Tool call blocks include get_info, and tool paths under home are abbreviated with a tilde.

This update improves file handling consistency across platforms by ensuring the effective search scope is visible and that current directories appear immediately upon opening.

Importance 1/3 Trust 1/3 llama.cpp Code generation Local inference

Read original

media Hugging Face Forums · 13d ago · 2 views

Forge generates and compiles sandboxed UIs for Hugging Face models in the browser

Amin Musah built Forge, an open-source browser playground for Hugging Face models that dynamically generates user interfaces using a separate code model. The system creates a TSX component based on task specs, which the browser then compiles client-side with esbuild-wasm and runs in a sandboxed iframe.

lab Google — The Keyword (AI) · 1h ago Live

Google and Kaggle host 353,000-person vibe coding course on AI agents

Google and Kaggle collaborated on the "5-Day AI Agents: Intensive Vibe Coding Course," which attracted over 353,000 registered participants. The free, multimodal course focused on teaching developers to program through natural language and deploy production-grade AI agents in the cloud.

github llama.cpp · 3h ago

llama.cpp b10270 adds Qwen3-TTS support with breaking llama-tts changes

The llama.cpp project released version b10270, introducing support for the Qwen3-TTS model. This update includes a breaking change to the llama-tts binary and implements significant architectural modifications to handle the new text-to-speech workflow.

media TLDR AI · 6h ago

ChronicleBio uses AI to cure POTS; Mind Lab tests Macaron-V1 continual learning

ChronicleBio, founded by former OpenAI exec Fidji Simo, is using AI and 153 terabytes of blood data to cure Postural Orthostatic Tachycardia Syndrome (POTS). Meanwhile, Mind Lab's Macaron-V1 model surpasses GLM-5.2 in benchmarks by dynamically switching between five LoRA expert modules for continual learning.

github llama.cpp · 8h ago

llama.cpp b10259 allows tensor reshape during load

The llama.cpp project has released build b10259, which introduces the ability to allow the reshape of tensors during the loading process.