Article URL: https://blog.google/innovation-and-ai/technology/developers-tools/embeddinggemma-2/ Comments URL: https://news.ycombinator.com/item?id=49980487 Points: 15 # Comments: 1
This is an AI-generated summary. ShortSingh links to the original source for the complete article.
Running a 180B-Parameter MoE Model on a Gaming Laptop: VIDRAFT's POCKET-Darwin-180B-GGUF TL;DR: VIDRAFT has released POCKET-Darwin-180B-GGUF, a 4-bit quantized, GGUF-format version of their 180B Mixture-of-Experts model that runs on consumer hardware — including a gaming laptop with 8 GB VRAM and 32 GB RAM. By combining MoE sparse activation with llama.cpp-based SSD streaming, only ~3B parameters are computed per token at inference time, slashing hardware requirements by roughly 250× compared to the full-precision server setup. Developers working on air-gapped or on-device deployments should p
If SSH listens on a nonstandard port, SCP needs to connect to that same port. The easy-to-miss detail is capitalization: use uppercase -P with scp, not lowercase -p. SCP doesn't have a separate network port. It connects through SSH, which uses TCP port 22 by default. If a server's SSH service listens on port 2222, SCP must connect to 2222 too.
In enterprise core banking modernizations, microservice decomposition often creates an unexpected governance failure: Semantic Drift. When dozens of distributed product teams build microservices independently, subtle discrepancies emerge between architectural specifications and real-world implementations. A field defined as ISO4217 currency string in a central domain model might be implemented as a raw string in one service, a numeric currency code in another, or omitted entirely in an asynchronous event payload. Relying on a "Code-First" development approach—where engineers write Java control
What You’ll Build By the end of this post you’ll have a tiny but powerful “Podcast Intro Generator” that: Accepts a podcast title, host name, and a short tagline. Dynamically stitches those strings together into a polished intro script. Sends that script to a text‑to‑speech engine. Returns a high‑quality audio clip you can drop straight into your podcast editing workflow. We’ll use ElevenLabs for the heavy lifting—its API gives us instant, human‑like voices and even voice‑cloning capabilities.
Discussion (0)
Log in to join the discussion and vote.
Log in