Text-to-Speech
audio workload
Synthesizing speech from text. Latency-sensitive for interactive use; models are typically small relative to LLMs.
Classification
- Category
- audio
- Compute profile
- mixed
- Scaling
- horizontal
Infrastructure considerations
Qualitative — no fabricated numbers- Interconnect sensitivity
- low
- Network sensitivity
- low
Curated · Memory · Small-to-mid model weights + activations; latency-driven rather than memory-driven.
Curated · Storage · Model weights plus audio output storage.
Typical frameworks
Curated — not a catalog relationshipPyTorchHugging Face
Data class
Curated · The workload taxonomy is authored qualitative reference, not an ingested source. Sensitivities and profiles are classifications, not measurements; a workload's concrete VRAM comes from the specific model it runs.