You are working on the stanfordnlp/dspy repository. The source code is checked out at /testbed at commit…
You are working on the stanfordnlp/dspy repository. The source code is checked out at /testbed at commit…: a task in SWE-Lego-RL-2699 (Harbor dataset). The gpt-5-chat model (available through Azure AI Foundry) is being incorrectly classified as a reasoning model in DSPy, causing it to be subject…
The task
The `gpt-5-chat` model (available through Azure AI Foundry) is being incorrectly classified as a reasoning model in DSPy, causing it to be subject to reasoning model validation requirements (temperature=1.0 and max_tokens>=16000) that don't apply to this model.
Part of Lego-X/Lego-RL-2699.