True when a safetensors file's header names any audio_tower.* tensor
(header JSON only — never touches tensor data). Separates a real
audio-capable sidecar (e4b: 752 audio tensors) from the 12B stub, whose
only audio entry is embed_audio.embedding_projection.weight. Missing or
unparseable files read as false.
True when a safetensors file's header names any
audio_tower.*tensor (header JSON only — never touches tensor data). Separates a real audio-capable sidecar (e4b: 752 audio tensors) from the 12B stub, whose only audio entry is embed_audio.embedding_projection.weight. Missing or unparseable files read as false.