Abstract: This document relates to using machine learning models to automate the process of generating alert rules for computing systems. The disclosed techniques can generate templates that match log entries from system logs. Then, the templates can be employed to generate alert rules that specify certain conditions when alerts should be triggered. In some implementations, the machine learning model is invoked responsive to a determination that a given log entry does not already match an existing template. Thus, the number of invocations of the machine learning model is proportional to the number of templates, rather than the number of log entries. Since there are often many more log entries than templates, this approach can yield significant computational savings compared to approaches that might process each log entry using the machine learning model.

Abstract: Systems and methods for dynamically modifying a cache associated with a neural network model of a natural language generator are described. In examples, a neural network model employs a beam search algorithm at a decoder when decoding output and generating predicted output candidates. The decoder utilizes caching techniques to improve a speed at which the neural network operations. When an amount of memory utilized by one or more caches of the neural network model is determined to exceed a threshold memory size, a layer-specific portion of a cache associated with a layer of the neural network model is identified. The identified layer-specific portion of the cache can be deleted when the amount of memory utilized by the cache of the neural network model exceeds the threshold memory size. In examples, data in the cache is deduplicated and/or deleted.

Abstract: A method for charging a battery, comprising: providing a battery; coupling the battery to a charge and discharge device and to a battery management system; and with the battery management system, alternating cycles of charge and discharge of the battery until the battery is charged to a predetermined level, wherein each charge and discharge cycle comprises a charge of the battery at a high charge rate and a discharge of the battery at a low discharge rate; wherein the low discharge rate is smaller than or equal to 10% of the high charge rate.

Abstract: Methods and apparatuses for improving the performance and energy efficiency of Radio Access Networks (RANs) are described. Various power control schemes may dynamically adjust RAN power consumption based on fluctuations in network traffic, throughput, latency, queue sizes, and/or packet error rates with the goal of increasing energy efficiency while maintaining quality of service metrics. The power control schemes may be implemented using a PRB controller for dynamically allocating physical resource blocks (PRBs) to user devices and a CPU controller for assigning CPU power profiles based on PRB allocations for the user devices. The PRB controller and CPU controller may periodically acquire real-time telemetry data and wireless network performance information and then adjust the number of PRBs for user devices and adjust the CPU power profiles for executing RAN functions based on the telemetry data and wireless network performance information.

Abstract: A bi-level machine learning (“ML”)-assisted framework for resource management is provided that includes an application-level controller that operates in concert with a service-level controller to scale the allocation of a computing resource to a network service in a manner that satisfies a service-level objective (“SLO”) . The application-level controller utilizes a ML model to learn a performance target that is transmitted to the service-level controller. The service-level controller adjusts the allocation of the computing resource to the network service based on the performance target and resource utilization data associated with the network service.

Abstract: A vehicle, belonging to the field of battery technologies is described. The vehicle includes: a traction battery; an electric device; a charging interface, configured to connect to a charging device to enable the vehicle to be powered by the charging device, where the charging interface includes a positive electrode terminal and a negative electrode terminal; and a bypass module, configured to connect or disconnect the charging device from the electric device.

Abstract: A charging and heating apparatus includes a power unit configured to be connected to a battery to be charged/discharged and perform voltage conversion on a received voltage and current and output a converted voltage to the battery, a unidirectional/bidirectional conversion unit connected to the power unit and configured to output a bidirectional alternating current or a preset current to the power unit, and a control unit connected to the unidirectional/bidirectional conversion unit and configured to, in response to the battery being in a heating mode, control the unidirectional/bidirectional conversion unit to output the bidirectional alternating current to the corresponding power unit, such that the battery self-heats under excitation of the bidirectional alternating current, and in response to the battery being in a charge/discharge mode, control the unidirectional/bidirectional conversion unit to output the preset current to the power unit, such that the battery is charged/discharged.

Abstract: A housing for an electronic device can include an exterior titanium portion, an interior metal joined to the exterior titanium portion, the interior metal being a different metal than the exterior titanium portion, and an intermetallic compound having a thickness of less than 1 ?m disposed between the interior metal and the exterior titanium portion.

Abstract: A battery heating apparatus includes a voltage transformation module and a control module. The voltage transformation module includes a first voltage transformation circuit configured to connect a charging apparatus, a first switch module connected to the first voltage transformation circuit, a second voltage transformation circuit configured to connect a power battery, and an energy storage element and a second switch module connected to the second voltage transformation circuit. The control module is configured to, in a first heating mode, control the first switch module to turn off the first voltage transformation circuit, and control the second switch module so that the second voltage transformation circuit and the energy storage element form a loop for discharging from the power battery to the energy storage element and a loop for charging the power battery by the energy storage element, for heating the power battery during discharging and charging.

Abstract: Described are examples for detecting anomalies in files. Each of multiple files can be processed to generate corresponding flat files. Values can be extracted from multiple lines of each of the flat files into a parameter vector. Patterns can be generated from the multiple lines based on the values extracted. The parameter vector and the patterns can be provided as input to a machine learning (ML) model to obtain a set of rules for the files based on conditional probabilities. The set of rules can be applied to a set of one or more files to detect anomalies in values in the set of one or more files.

Abstract: Embodiments of this application provide a DC/DC conversion circuit, a power unit, a charging pile, and a charge-discharge heating method. The circuit includes: a first rectifier module, where an input end of the first rectifier module is connected to a power grid through an AC/DC conversion circuit; a transformer module, where an input end of the transformer module is connected to an output end of the first rectifier module; an energy storage module; and a second rectifier module, where an input end of the second rectifier module is configured to connect to an output end of the transformer module or the energy storage module, and an output end of the second rectifier module is configured to connect to a battery pack of an electric vehicle when the charging pile is charging the electric vehicle.

Abstract: Methods and apparatuses for improving the performance and energy efficiency of Radio Access Networks (RANs) are described. Various power control schemes may dynamically adjust RAN power consumption based on fluctuations in network traffic, throughput, latency, queue sizes, and/or packet error rates with the goal of increasing energy efficiency while maintaining quality of service metrics. The power control schemes may be implemented using a PRB controller for dynamically allocating physical resource blocks (PRBs) to user devices and a CPU controller for assigning CPU power profiles based on PRB allocations for the user devices. The PRB controller and CPU controller may periodically acquire real-time telemetry data and wireless network performance information and then adjust the number of PRBs for user devices and adjust the CPU power profiles for executing RAN functions based on the telemetry data and wireless network performance information.

Abstract: Computing technology is described herein that provides an attention mechanism, implemented by a neural network, that generates attention information based on head-specific query information and shared key and value (KV) information, without computing head-specific key information and head-specific value information, and without caching the head-specific key information and the head-specific value information in memory. This manner of operation allows the computing technology to make efficient use of processing and memory resources. In some implementations, the attention mechanism is part of decoder of an encoder-decoder system, or a standalone decoder system. In some implementations, the computing technology leverages the attention information to generate synthesized text based on input text.

Abstract: A motor controller includes: a plurality of first power modules, a second power module, a first group of connecting interfaces, a second group of connecting interfaces, and a first capacitor, where the plurality of first power modules are connected in parallel to each other, and the plurality of first power modules are configured to connect to the first group of connecting interfaces; the second power module is configured to connect to the second group of connecting interfaces; and the first capacitor is connected in parallel to the second power module, where the plurality of first power modules and the second power module are configured to perform temperature regulation on an object connected to the first group of connecting interfaces and the second group of connecting interfaces.

Abstract: Systems and methods for dynamically modifying a cache associated with a neural network model of a natural language generator are described. In examples, a neural network model employs a beam search algorithm at a decoder when decoding output and generating predicted output candidates. The decoder utilizes caching techniques to improve a speed at which the neural network operations. When an amount of memory utilized by one or more caches of the neural network model is determined to exceed a threshold memory size, a layer-specific portion of a cache associated with a layer of the neural network model is identified. The identified layer-specific portion of the cache can be deleted when the amount of memory utilized by the cache of the neural network model exceeds the threshold memory size. In examples, data in the cache is deduplicated and/or deleted.

Abstract: A battery main box, a charge-discharge circuit, and a powered device. The battery main box includes two first electrode circuits and one second electrode circuit. A switch is connected between the two first electrode circuits. The two first electrode circuits are respectively configured to be connected to two powered loads and two batteries, and the second electrode circuit is configured to be respectively connected to the two powered loads and the two batteries. The battery main box in the embodiments of the present application includes the two first electrode circuits and the one second electrode circuit, and the switch is connected between the two first electrode circuits.

Abstract: A charge and discharge control method comprises: when a battery heating condition is satisfied, turning on a connecting circuit between a neutral point of a first electric motor and a neutral point of a second electric motor, wherein a first electric motor controller connected to the first electric motor and a second electric motor controller connected to the second electric motor are both connected between a positive electrode and a negative electrode of a battery; and controlling the turning-on/turning-off of bridge arms in the first electric motor controller and the second electric motor controller, and forming, in a circuit in which the battery is located, a charging loop and a discharging loop which are alternately switched.

Abstract: Embodiments of the present application provide a charging and discharging apparatus and a battery charging method, which are capable of ensuring security performance of a battery. The apparatus comprises a bi-directional DC/DC converter and a control unit, wherein the control unit is configured to: receive a first charging current transmitted by a battery management system (BMS) of a battery, control the bi-directional DC/DC converter based on the first charging current to charge the battery through an energy storage battery; receive a first discharging current transmitted by the BMS and control the bi-directional DC/DC converter based on the first discharging current to discharge a battery capacity of the battery to the energy storage battery; and receive a second charging current transmitted by the BMS and control the bi-directional DC/DC converter based on the second charging current to charge the battery through the energy storage battery.

Abstract: A charge-discharge circuit and an electric device. The charge-discharge circuit includes a power supply module, a first drive assembly, a second drive assembly, and a first switch module, where the first drive assembly and the second drive assembly are connected in parallel between a positive terminal and a negative terminal of the power supply module; and a terminal of the first switch module is connected to the first drive assembly, and another terminal of the first switch module is connected to the second drive assembly.

Abstract: A charging and discharging circuit, a battery control circuit and an electrical device. The charging and discharging circuit includes: a power supply module, a first driving assembly, a second driving assembly and a first switch; the power supply module includes a battery, the first driving assembly includes a first motor, and the second driving assembly includes a second motor; the first driving assembly and the second driving assembly are connected in parallel between a positive electrode end and a negative electrode end of the power supply module; and one end of the first switch is connected to a neutral point of the first motor, and the other end of the first switch is connected to a neutral point of the second motor.