标签：#ha

We are so excited to announce the winners of the Gemma 4 Challenge! This is officially our most...

VS Code 1.123 adds a two-hour delay before auto-updating extensions to newly published versions, creating a revocation window against supply chain attacks. The delay does not apply to trusted publishers like Microsoft, GitHub, and OpenAI. Similar cooldown mechanisms have now spread across pip, RubyGems, npm, pnpm, Yarn, and Bun. By Steef-Jan Wiggers

Solar geoengineering is often portrayed as a sort of emergency brake. Something along the lines of Pull in case of climate emergency to scatter light-reflecting particles to bounce sunlight out of the atmosphere and cool the planet. But it might be less like a simple brake and more like a complicated, entirely unsolved puzzle. Some…

What flipped in b9437 Build b9437 , published on May 30, 2026 at 20:56 UTC , ships two targeted default-value corrections to llama-bench . Flash attention ( -fa ) shifts from a hard-coded off to auto ( LLAMA_FLASH_ATTN_TYPE_AUTO ), and the GPU-layer count ( -ngl ) changes from the legacy sentinel 99 to -1 . Both values now match what llama-server and llama-cli already used — the bench tool was simply never updated to track them until this build. Quick Answer: Before b9437 (published May 30, 2026) , llama-bench hard-coded -fa off , silently skipping flash attention even on CUDA, Metal, and Vulkan hardware. Build b9437 sets the default to -fa auto and -ngl -1 , matching llama-server and llama-cli . Any pre-b9437 baseline on FA-capable hardware needs a flag-matched re-run to remain valid. PR #23714 , reviewed and merged by maintainers JohannesGaessler and pwilkin, adds the same -fa auto|off|on tri-state flag to llama-bench that the rest of the toolchain already supported. With LLAMA_FLASH_ATTN_TYPE_AUTO as the new default, flash attention activates automatically when the runtime detects a capable backend (CUDA, Metal, Vulkan); on CPU-only hosts it stays off with no error and no output change. Parameter Before b9437 After b9437 Behavioral impact -fa off (hard-coded) auto ( LLAMA_FLASH_ATTN_TYPE_AUTO ) GPU-capable hosts bench with FA active by default; pre/post comparisons require explicit flag-matching -ngl 99 (offload-all sentinel) -1 (runtime decides) CPU-only builds no longer attempt full GPU offload; eliminates spurious CUDA errors when no GPU is present The following verified script (executed successfully, exit 0) demonstrates the behavioral gap in concrete terms — on a capable GPU, the pre-b9437 defaults schedule zero FA rows while b9437 defaults schedule one: def old_llama_bench ( device ): # Before b9437, the default bench matrix used FA=0, so FA rows were skipped. return [{ " device " : device [ " name " ], " ngl " : 0 , " fa " : 0 }] def b9437_llama_bench ( de

The A2A protocol is an agent to agent protocol from Google. I was looking at it 6 months back but it didn't feel like I really understood how to use it at that point. Probably because we were all still trying to figure out agents and then the MCP protocol became quite a big deal. But now I'm starting to think that once an agent has tools and services and data and then contacts. Actually, the point of interaction becomes the agent itself and then if you build other agents you would want them to i