Allow no-timestamp Whisper decoding beyond 224 tokens

Problem

WhisperReplica::generate currently limits decoding with:

decoding_options.max_length = std::min(total_max_length / 2, total_max_length - start_step);

This applies the original 224-token sampling limit even when the prompt ends in <|notimestamps|>. For languages that rely heavily on Whisper's byte-level fallback, 224 tokens can be exhausted by a short audio clip even though the model has a 448-position text context.

Reproduction

Using faster-whisper with large-v3, vad_filter=False, and without_timestamps=True, both an Armenian and a Georgian sample stopped at exactly 224 re-tokenized visible-text tokens. One output ended with the Unicode replacement character because generation stopped mid-byte sequence.

The same audio with timestamp decoding used more than 224 visible-text tokens because timestamp boundaries allow faster-whisper to continue the transcription loop, but disabling timestamps produces one decode for the complete audio window.

Proposed behavior

When the task prompt ends with <|notimestamps|>, allow decoding to use the remaining text context:

total_max_length - start_step

Keep the existing half-context limit for timestamped decoding. This preserves the standard timestamp path while allowing explicit no-timestamp callers to use the model's remaining decoder positions.

Start at WhisperReplica::generate and inspect how decoding_options.max_length is computed from the task prompt and start_step. Reproduce with no-timestamp decoding and verify that prompts ending in <|notimestamps|> can use the remaining text context, while timestamped decoding keeps the existing half-context limit.