Found 1 document(s)
Page 1 of 1
-
Edge-cloud split inference runs the early layers of a multimodal large language model (MLLM) on a device and sends intermediate features to a server to complete the task. While keeping raw images local is often perceived...