Abstract: Modern automatic speech recognition (ASR) systems can utilize artificial intelligence (AI) models to service ASR requests. The number and scale of AI models used in a modern ASR system can be substantial. The process of configuring and reconfiguring hardware to execute various AI models corresponding to a substantial number of ASR requests can be time consuming and inefficient. Among other features, the described technology utilizes batching of ASR requests, splitting of the ASR requests, and/or parallel processing to efficiently use hardware tasked with executing AI models corresponding to ASR requests. In one embodiment, the compute graphs of ASR tasks are used to batch the ASR requests. The corresponding AI models of each batch can be loaded into hardware, and batches can be processed in parallel. In some embodiments, the ASR requests are split, batched, and processed in parallel.

Abstract: A device may receive a create session (CS) request associated with establishing a packet data network (PDN) connection with a user device. The device may determine, based on receiving the CS request, that the CS request is to be redirected for establishment of the PDN connection. The device may identify, based on determining that the CS request is to be redirected, redirect information that identifies a particular PDN gateway to which the CS request is to be redirected. The device may provide a CS response, including the redirect information, to permit the PDN connection to be established between the user device and the particular PDN gateway.

Abstract: A device may receive a create session (CS) request associated with establishing a packet data network (PDN) connection with a user device. The device may determine, based on receiving the CS request, that the CS request is to be redirected for establishment of the PDN connection. The device may identify, based on determining that the CS request is to be redirected, redirect information that identifies a particular PDN gateway to which the CS request is to be redirected. The device may provide a CS response, including the redirect information, to permit the PDN connection to be established between the user device and the particular PDN gateway.

Abstract: An embodiment of the invention provides a method wherein a database query including a first constraint and additional constraint(s) are received in a first node. Data in the first node that satisfies the first constraint is identified, encoded, and sent to a second node. Encoded data is identified in a mapping table in the second node; and, one or more missing identifiers are identified that include encoded data that is not in the mapping table. The missing identifier is sent to the first node, decoded to retrieve the value of the missing identifier, and mapped to the retrieved value. The mapping of the missing identifier and the retrieved value are sent to the second node. A dictionary in the second node is queried with the retrieved value to identify an identification number for the retrieved value. The missing identifier is mapped to the identification number for the retrieved value.

Abstract: Disclosed is a solid, floating soap which has a formulation density greater than water or a liquid of like density, and which has an overall physical density less than water or a liquid of like density, and a method of making the floating soap.

Abstract: Disclosed is a solid, floating soap which has a formulation density greater than water or a liquid of like density, and which has an overall physical density less than water or a liquid of like density, and a method of making the floating soap.