Inference profiles now accept provider-native model identifiers

When configuring an inference profile in the Hive dashboard, the model field now accepts the identifier the upstream provider actually expects, for example accounts/fireworks/models/minimax-m3 for Fireworks or gpt-4o-mini for OpenAI. Previously the form required a provider-prefixed value and rejected valid Fireworks paths, so operators could not save those profiles. Existing saved profiles that still carry the old prefix continue to work because Hive now strips the prefix only when it is present.