I think I have seen enough about DeepSeek models and research now to be confident that there is indeed innovation coming from them. I am especially interested in the domain of reinforcement learning here because it seems to me that they took an important leap with GRPO just like OpenAI did with RLHF
31.01.2025 21:24 — 👍 3 🔁 1 💬 0 📌 0I wonder what it’s good for. Probably not something I need to solve a lot 🤣🤷♂️
31.01.2025 21:20 — 👍 1 🔁 0 💬 0 📌 0Who here has tried pydantics new agent framework? I am pretty excited about it since i think that we finally have an interface that looks promising and that has the focus on developer experience we need. Also, it looks like it is solves the issue with being able to interact between runs.
04.12.2024 19:43 — 👍 4 🔁 0 💬 0 📌 0What is Theis place?
21.11.2024 18:24 — 👍 3 🔁 0 💬 0 📌 1