Nadia

Worth checking whether the runtime is applying the template at all, because there are two layers here and they can disagree.

If you are calling a chat-style endpoint, the server applies the template for you. If you are calling a raw completion endpoint, it does not — you are responsible for the delimiters, and passing a bare user message means no template, which produces exactly this.

A lot of people switch endpoints while debugging something else, and inherit this problem without connecting the two.

Most servers will log the fully-templated prompt if you turn the verbosity up. Look at it once. It is usually immediately obvious whether the delimiters are there.

For the immediate unblocking, before the proper fix: check whether your CI provider has a status page and whether anyone else is reporting the same thing.

A meaningful share of "broke with no changes" incidents are the provider's, or a major registry's, and they are resolved within hours by someone else. Ten seconds of looking saves you from a deep investigation into a problem that is not yours.

Same instinct for the ecosystem: a broken release of a very widely used package produces a lot of noise very quickly, and there is usually already a thread about it.

Symptom worth adding to the list, because it is the quiet version of the same bug: subtly worse quality rather than obvious garbage.

If padding is only slightly wrong — a couple of pad tokens, or position ids off by the pad count — you do not get nonsense. You get answers that are a bit worse, a bit more repetitive, occasionally off-topic. In an evaluation that reads as the model being mediocre, and people spend days on prompts and sampling parameters chasing it.

Rule of thumb: if batch size one gives noticeably better output than batched inference on the same prompt, stop tuning anything else and look at your padding. Same prompt, same seed, both ways — that comparison finds it in a minute.