Excited to launch Gemma 4: the best open models in the world for their respective sizes. Available in 4 sizes that can be fine-tuned for your specific task: 31B dense for great raw performance, 26B MoE for low latency, and effective 2B & 4B for edge device use - happy building!
llmfit 好讚,直接預估,輸出速度跟記憶體用量,有 TUI 跟 Web 介面,
不過我懷疑他的 token per second 太樂觀,我在 too tight 的狀況下,用 lmstudio 大多 < 1 token per second。
github.com/AlexsJones/llm…
LLM 神經解剖學 - 如何在不改變任何權重的情況下榮登 LLM 排行榜榜首 by David Noel Ng
這是 Twinkle AI "發呆貓" 轉貼的文章,不知道大家的看法如何 ?
作者發現:不需要重新訓練模型、也不需要改權重,只要把模型中間一小段 layer「再跑一次」,有時就能讓模型表現變好。
Anthropic published a blog post one hour ago.
Cybersecurity stocks have lost $10B since.
CrowdStrike -6.5%. Cloudflare -6%. Okta -5.7%.
One blog post. One hour. $10B gone.