「立派な科学者ですね」 校内放送で褒められた自由研究がすごすぎる 12万人が驚き | Hint-Pot hint-pot.jp/archives/185639
posted at 06:23:17
RT @DrJimFan: DeepMind's paper turns textual LLMs into general-purpose compressor API, even for images & audio. Now it can be compared directly to gzip, PNG, LZMA2, etc. We already know that learning & generalization are closely linked to compression, but it's fascinating to see this… twitter.com/i/web/status/1… pic.twitter.com/HzH0f6mryM
posted at 04:37:22
RT @akshay_pachaar: Everyone should learn how to fine-tune LLMs. However, LLMs barely fit in GPU memory❗️ This is where fine-tuning & more importantly parameter efficient fine-tuning (PEFT) become essential. Let's understand them today: pic.twitter.com/T93uv7vu2W
posted at 04:24:28
RT @osanseviero: Lots of people are interested in model merging! 👀 Here is a collection of some of the most important papers in the area + some minor notes. Any I should add? huggingface.co/collections/os…
posted at 04:16:40
RT @llama_index: Over the past week we’ve built an entire tutorial series on “building RAG from scratch” - supercharge your AI skills and learn how things work under the hood! 🚀 Ingestion, vector db, retrieval, LLM synthesis, routing, evals. More coming soon 🔥 gpt-index.readthedocs.io/en/latest/end_… pic.twitter.com/I34lQSSFp8
posted at 04:15:08
RT @studiomasakaki: 前回の線画をAIイラストにする実験に引き続き、私が塗り塗りしたトンチキなカラーラフをAIに仕上げてもらう検証をやってみました。X/Y plotで総当りした「tile」の最適パラメータも紹介しています。 カラーラフを『雑塗りtile』！適正パラメータ総当り実験studiomasakaki.fanbox.cc/posts/6706815 twitter.com/i/web/status/1… pic.twitter.com/x46ipEc1se
posted at 04:14:34
RT @jure: 🌟 Excited to announce the Stanford Graph Learning Workshop 2023 on Oct 24 2023!🤝 Bringing together academia & industry leaders to delve into advances in #MachineLearning & #AI in Relational domains, Foundation models, and Multimodal AI. 📢 Free registration. 📢 Calling for… twitter.com/i/web/status/1… pic.twitter.com/RjFzh1scPH
posted at 04:13:12
RT @jerryjliu0: There’s an entire basket of techniques that you can try to improve your RAG system 🧺: ✅ chunk sizes ✅ fine-tuning embeddings ✅ advanced retrieval And way more, all included in @llama_index. But how do you know what works and what doesn’t? 💡You need to have evaluation… twitter.com/i/web/status/1… pic.twitter.com/YbNj9PAOvP
posted at 04:12:59