Tip for developers testing function calling: In these open models, tool calling is part of the chain-of-thought. When the model calls a tool, keep in mind you need to pass that reasoning back when sending the tool’s result.
CODE
-
LLM Apps Repository for Building AI Agents and RAG
By
–
Awesome LLM Apps Repo for building AI Agents and RAG Apps 55k+ stars and counting
-
GPT-OSS Demo: Local Laptop Execution with Tools and Game Coding
By
–
Watch @DKundel and @zhuohan123 bring gpt-oss to life: running locally on a laptop, calling tools, and coding a game entirely offline with the Codex CLI. Fun to be behind the scenes for this one as the demos came together on Sunday!
-
Graph Breaks in AI Training: Python Runtime Considerations
By
–
Graph breaks are a fact of life, and that's fine. You can't really expect your entire end to end training to never touch the python runtime. It doesn't mean that you should try to somehow reinvent Python compilation from scratch however.
-
PyTorch Team Shows Efficient Llama Training Methods
By
–
Yes I know. We helped show the PyTorch team how to train Llama efficiently – they have a whole GPU Mode talk about it. These issues do not at all line with your claims however.
-
Claude Opus 4.1 upgrade boosts AI coding reliability
By
–
3/ Claude Opus 4.1 is an upgrade of already the most powerful coding model. This improvement is important for vibe coders, making their code more reliable!
-
Tool Calling Implementation Challenges in New Harmony Models
By
–
Sounds to me like it's tool calling, which is the bit of the model I've not got working yet myself – the new Harmony stuff means there are plenty of details that model serving tools need to nail down before I'll be confident if we know how well that actually works
-
Open Source AI Model Training Details and Documentation
By
–
Added an extra section with interesting details from the model card about how the models were trained https://
simonwillison.net/2025/Aug/5/gpt
-oss/#the-model-card
… -
GPT-OSS Models Now Available in AI SDK Playground
By
–
Both gpt-oss-120b and gpt-oss-20b are now in @aisdk and you can test them nicely side-by-side in the AI SDK playground. ▲
-
Training Methods Critical for AI Model Performance
By
–
well they have to train it like this or it wouldn't work nearly as well