Abstract: This application discloses a packet transmission method, an apparatus, a device, and a readable storage medium, and relates to the field of communication technologies. The method applied to a second network device includes: First, a first packet sent by a first network device is received, and then a second packet, a first dwell time period, and a second dwell time period are obtained based on the first packet. Then, a time difference between the second dwell time period and the first dwell time period is determined. Then, a third dwell time period and a fourth dwell time period are determined, to encapsulate the third dwell time period, the fourth dwell time period, and the second packet to obtain a third packet.

Abstract: Embodiments described herein provide a system for selecting a neural network based natural language processing (NLP) model for building a custom artificial intelligence (AI) stack for a user. The system includes a communication interface that established connections to one or more external servers hosting one or more neural network based NLP models, a memory; and a processor executing operations including: selecting a source document based on a custom NLP application; generating, by a first language model, a summary of the source document; generating, by a second language model, one or more questions based on at least one of the summary or the source document; transmitting, via the communication interface, the one or more questions to the one or more neural network based NLP models; receiving, via the communication interface, one or more answers generated by the one or more neural network based NLP models.

Abstract: Embodiments described herein provide a framework that integrates a retriever model and the LLM to feed retrieved passages to an LLM to generate an answer conditioned on the retrieved passages in response to a query. For example, in one embodiment, a single-round approach is implemented, which involves directly transmitting the retrieved passages to the LLM. For another example, a multi-round methodology is implemented, which involves initially presenting the retrieved passages to the Language Model, collecting its responses, and then adjusting our interaction with the Language Model based on this acquired feedback.

Abstract: Embodiments described herein provide a semantic parsing framework which may be referred to as Uni-Parser. The Uni-Parser framework may be applied to question answering on both knowledge bases and databases. The three main stages of the Uni-Parser framework are enumeration, ranking, and generation. At the enumeration stage, primitives are enumerated based on matching the question to the data structure. After enumerating primitives, the Uni-Parser framework may rank the primitives used a trained ranker model. The top ranked primitives may then be used as inputs to a generator which is a learned sequence to sequence model which produces a logical form.

Abstract: A hand-held steam cleaning apparatus including: a housing, provided with a handle; a steam generating unit, including a heating body and configured to heat water and vaporize the water into steam; a steam nozzle, configured to output the steam; and a water supply device, including a water tank and pump configured to pump the water in the water tank into the steam generating unit. The cleaning apparatus is powered by a battery pack. The weight of the cleaning apparatus is not greater than 2 kg when the housing is fitted with the battery pack and the water tank is fully loaded; the capacity of the battery pack is in the range of 36-144 Wh. The steam flow capacity is in the range of 3-12 g/min; and the continuous work time is not less than 2 min.

Abstract: A packet transmission method and apparatus, and a system. The method includes: after receiving a first packet, a first device determines, based on a condition met by a first number carried in the first packet, a manner of sending the first packet. When the first number of the first packet meets a first condition, a second device determines a second cycle label corresponding to a first cycle label and sends the first packet to a third device in a cycle corresponding to the second cycle label. When the first number of the first packet meets a second condition, the second device determines a third cycle label corresponding to the first cycle label and sends the first packet to the third device in a cycle corresponding to the third cycle label.

Abstract: Embodiments described herein provide systems and methods for question answering using a hybrid question parser and executor model. The hybrid question parser and executor model includes a hybrid parser model and a hybrid executor model. The hybrid parser model includes a first neural network model, and generates a representation of an input question. The representation includes primitives and operations representing relationships among the primitives. The hybrid executor model generates an answer to the input question by executing the representation based on an input text document. The hybrid executor model includes an execution neural network model for executing the primitives of the representation, and an execution programming model for executing the operations of the representation.

Abstract: Embodiments described herein provide systems and methods for training a text retrieval model. A system may generate queries associated with provided documents. The queries may be generated in one or more different manners. Examples of query generation may include extracting relevant spans of text from the documents, prompting a language model for a topic, title, abstractive summary, and/or extractive summary based on the documents. Metadata such as title or other HTML tags may be used as queries. Using the one or more queries, the text retrieval model may be trained using contrastive learning, using the generated query, and positive and negative sample documents. A fine-tuning training phase may be performed using domain-specific data which may also be done with generated query pairs, or may be done in a supervised fashion with provided queries. The text retrieval model may be used to locate documents given an input query.

Abstract: This application provides a Example communication methods and apparatus, to improve packet scalability are described. In one example, a method includes: A first node receives a first packet from a second node, and parses the first packet. The first packet includes multicast routing information of the second node, and the first node is a child node of the second node in a multicast tree. A multicast routing identifier of a node is used by a non-leaf child node of the node in the multicast tree to determine multicast routing information of the non-leaf child node, and multicast routing information of a node includes a multicast routing identifier of the node and multicast routing information of a non-leaf child node of the node in the multicast tree.

Abstract: Embodiments described herein provide a semantic parsing framework which may be referred to as Uni-Parser. The Uni-Parser framework may be applied to question answering on both knowledge bases and databases. The three main stages of the Uni-Parser framework are enumeration, ranking, and generation. At the enumeration stage, primitives are enumerated based on matching the question to the data structure. After enumerating primitives, the Uni-Parser framework may rank the primitives used a trained ranker model. The top ranked primitives may then be used as inputs to a generator which is a learned sequence to sequence model which produces a logical form.

Abstract: Embodiments described herein provide systems and methods for multimodal layout generations for digital publications. The system may receive as inputs, a background image, one or more foreground texts, and one or more foreground images. Feature representations of the background image may be generated. The foreground inputs may be input to a layout generator which has cross attention to the background image feature representations in order to generate a layout comprising of bounding box parameters for each input item. A composite layout may be generated based on the inputs and generated bounding boxes. The resulting composite layout may then be displayed on a user interface.

Abstract: This application provides a packet forwarding method and a related apparatus. The method may include: A first device receives a first packet sent by a second device in a second cycle; and the first device determines a target cycle corresponding to the first packet based on cycle modification information corresponding to the first packet and a cycle mapping relationship, and forwards the first packet in the target cycle corresponding to the first packet.

Abstract: A multicast packet sending method includes a first node that obtains a first multicast packet including multicast routing information of a second node, and sends the first multicast packet to the second node. The second node is a non-leaf subnode of the first node in a multicast tree. Multicast routing information of a node includes a multicast routing identifier of the node and multicast routing information of a non-leaf subnode of the node in the multicast tree. A multicast routing identifier of a node is used by the node to determine a subnode of the node in the multicast tree. The multicast routing information is embedded in the multicast packet, and multicast routing information of a tree recursive structure included in the multicast packet can describe packet forwarding information of the multicast tree or a subtree of the multicast tree.

Abstract: A packet processing method is disclosed. The method includes: A first network device receives a first packet from a network at a first moment, where the first packet is a 1st packet in a first burst of a first data flow, the first burst is one of a plurality of bursts included in the first data flow received by the first network device, the first burst includes one or more packets, and the first network device is a first-hop network device that processes one or more packets included in the first data flow. The first network device determines, based on the first moment, a first target queue from a plurality of queues included in a first queuing system of the first network device.

Abstract: Embodiments of the present invention provide a packet processing method, which is applied to a first node. The method includes: obtaining N first packets, where N is an integer greater than or equal to 1, and each of the first packets includes a first label; and sending the N first packets in a sending period corresponding to the first label.

Abstract: This application provides a traffic shaping method and apparatus. The method includes: A packet marking apparatus receives a first packet; the packet marking apparatus determines an enqueuing queue of the first packet; and the packet marking apparatus marks a queue identifier of the first packet as a queue identifier of the enqueuing queue of the first packet, and then sends the queue identifier of the first packet to a packet output apparatus, where the packet output apparatus is configured to send, based on the queue identifier of the first packet, the first packet to a corresponding queue for outputting. Therefore, packet output time after traffic shaping can be determined.

Abstract: This application provides data packet scheduling methods and apparatuses. One method includes: A first network device receives, at a first moment, a data packet from a second network device in a network, the first network device determines a first reference moment based on the first moment and time information carried in the data packet, the first network device determines, based on the first reference moment, a target queue from a plurality of queues included in a first queue system and adds the data packet to the target queue, and the first network device processes the target queue according to a scheduling rule of the plurality of queues.

Abstract: This application discloses a packet transmission method, an apparatus, a device, and a readable storage medium, and relates to the field of communication technologies. The method applied to a second network device includes: First, a first packet sent by a first network device is received, and then a second packet, a first dwell time period, and a second dwell time period are obtained based on the first packet. Then, a time difference between the second dwell time period and the first dwell time period is determined. Then, a third dwell time period and a fourth dwell time period are determined, to encapsulate the third dwell time period, the fourth dwell time period, and the second packet to obtain a third packet.

Abstract: A data processing method, a packet sending method, and an apparatus. The data processing method includes: a first device obtains a first periodicity parameter and a second periodicity parameter, where the first periodicity parameter is determined based on duration of transmitting a packet from a first network device to a second network device along a first path, and the second periodicity parameter is determined based on duration of transmitting a packet from the first network device to the second network device along a second path; and the first device determines a third periodicity parameter based on the first periodicity parameter and the second periodicity parameter, where the third periodicity parameter reflects a quantity of periodicities between a periodicity in which the first network device sends a data packet and a periodicity in which the second network device forwards the data packet.

Abstract: Embodiments of this application provide a method for sending a data packet. The method includes: A second network device may determine a remaining processing time of a data packet on a first network device based on a predefined first delay. When the remaining processing time is greater than 0, a moment at which the data packet enters a queue on the second network device serves as a start moment, and after a period of time, the remaining processing time ends at a second reference moment corresponding to the data packet. The remaining processing time is consumed on the second network device.