I have the same issue with gpt-oss-120b

I have the same issue with gpt-oss-120b: it doesn't follow the JSON format from time to time. I am unsure whether these models are quantized. or something else is going on
Was this page helpful?