Recent posts
Page 38 of 51 · 603 posts
Posted May 4
Daniel Lemire 对比了Amazon Graviton 3, Apple M2, Qualcomm 8cx 3rd gen 这三块 Arm CPU,结果 M2 优于 Qualcomm 8cx,优于 Graviton 3。 https://lemire.me/blog/2023/05/03/graviton-3-apple-m2-and-qualcomm-8cx-3rd-gen-a-url-parsing-benchmark/
Posted May 4
https://matt-rickard.com/the-python-family-of-languages
Posted Apr 29
终于等到了 GPT-4 API,还只能用 8K 的模型,32K 的用不了。
Posted Apr 23
同样的数据放到 blog 文章上,是不是也适用? https://twitter.com/robertoblake/status/1649785158294282240
Posted Apr 22
https://twitter.com/tibo_maker/status/1649729840369610754
Posted Apr 21
https://twitter.com/awilkinson/status/1649139599854440449
Posted Apr 20
How Replit trains Large Language Models (LLMs) using Databricks, Hugging Face, and MosaicML https://blog.replit.com/llm-training
Posted Apr 18
经朋友推荐,看到这个域名搜索引擎,非常不错。 https://domainr.com/ 比如你想买一个 telegram 的域名,它会把相近的候选项自动列举出来。比如 telegr.am, Tele.gram 等组合
Posted Apr 16
Posted Apr 14
OpenAI 的 Lilian Li 专门介绍了 prompt engineering https://t.me/tms_ur_way/2655
Posted Apr 14
一键安装 gpt4all 的方法来了 https://gpt4all.io/index.html
Posted Apr 14
前几天和朋友聊到微软推出的 visualchat,一看,好家伙,22个模型串起来,至少要四张卡才能跑起来。 https://github.com/microsoft/visual-chatgpt#gpu-memory-usage