Encodings#
Note
Multiprocess (MP) support for encoding (encoder / multimodal) caching is coming soon. The current in-process implementation is preserved in the Legacy section: Encoder caching and KV Caching for Multimodal Models with vLLM.
Note
Multiprocess (MP) support for encoding (encoder / multimodal) caching is coming soon. The current in-process implementation is preserved in the Legacy section: Encoder caching and KV Caching for Multimodal Models with vLLM.