Abstract: The present disclosure relates to the field of data processing. Provided is a distributed computing method, comprising: acquiring a data computing task; splitting the data computing task to obtain subtasks, deploying the subtasks to computing nodes, and configuring a parallel mode for each of the computing nodes in a distributed training universal frame; configuring a connection manner and a communication synchronization manner between the computing nodes; optimizing information synchronization efficiency for the computing nodes by using a gradient optimization algorithm or a non-gradient optimization algorithm; and aggregating intermediate results generated by the computing nodes, and outputting a corresponding final computing result.

Abstract: An image generating method and apparatus, and a device and a medium are disclosed. The method comprises: acquiring weakly correlated image-text data pairs, and creating an image-text data set according to the weakly correlated image-text data pairs, wherein the weakly correlated image-text data pairs are image-text data pairs in which images and texts have weak correlations (S11); training, by using the image-text data set, an image generation model which is preconstructed on the basis of an adversarial network, so as to obtain a trained image generation model, wherein the image generation model includes a generator for generating an image, and a discriminator for identifying the authenticity of the image and calculating a corresponding loss value (S12); and after when text data to be processed has been acquired, generating, by using the trained image generation model, an image corresponding to the said text data (S13).

Abstract: This application discloses a method of determining a network path and a network device. The method includes: determining a first reachable network path that uses a source node as a start point and a directly connected node of the source node as an end point; selecting a first node from a to-be-added node set to obtain a second reachable network path that uses the source node as the start point and the first node as a relay; and when there are m paths that use a same destination node as end points in the first and second reachable network paths, comparing the m paths based on n target factors to obtain a first dominant network path, and updating the first reachable network path based on the first dominant network path, where m?2, and n?2.

Abstract: A data sending method includes a first computer device predicting a link quality of N links based on a sending queue length of each of the N links and historical statistical information, and selecting, based on the predicted link quality, a first target link for sending first data, to send the first data to a second computer device, where the first data is a part of data in a target data stream.

Abstract: The technology of this application relates to a communication method and apparatus. The communication method includes a data sending device obtaining to-be-sent data of a locally installed first application, where a destination address of the to-be-sent data is an address of a data receiving device. The data sending device determines an address or addresses of one or more network adapters of the data receiving device based on the address of the data receiving device, the data sending device encapsulates the to-be-sent data based on the network adapter address to obtain encapsulated data, and the data sending device sends the encapsulated data to the data receiving device through a plurality of links. A destination address of each link is one of network adapter addresses of the data receiving device, and a source address of each link is one of network adapter addresses of the data sending device.

Abstract: A probiotic-encapsulating gum Arabic (GA) composite fiber/capsule, including: lactic acid bacteria and an electrospun fiber or electrosprayed capsule, where the lactic acid bacteria are encapsulated in the electrospun fiber or electrosprayed capsule by electro-hydro dynamics (EHD) technology. The electrospun fiber or electrosprayed capsule is prepared by compounding a polymer matrix with GA, and the polymer matrix is polyvinyl alcohol (PVOH), polyvinylpyrrolidone (PVP), whey protein concentrate (WPC) or maltodextrin (MD). A preparation method and an application of the composite fiber/capsule are also provided.

Abstract: A method and apparatus for allocating a computing task of a neural network in heterogeneous resources, a computer device, and a storage medium. The method includes: acquiring task information of the computing task of the neural network and resource information of the heterogeneous resources; determining, according to the task information and the resource information, an allocation mode for allocating each subtask to the heterogeneous resources for execution and a task processing cost corresponding to each allocation mode; constructing a directed acyclic graph according to each allocation mode and each task processing cost; obtaining a value of a loss function corresponding to each allocation path according to the task processing cost corresponding to each subtask in an allocation path of the directed acyclic graph; and selecting a target allocation path according to the value of each loss function.

Abstract: Systems and methods are provided for data shuffling for distributed machine learning training, including each training node in the network receiving a shard of training data, wherein the training data set is divided into shards having data items. Each data item is assigned to a working set such that each of the working set includes data items from multiple shards. The training nodes perform training using the data items of a first working set that are in each node's shard. Upon completion of the training using the data items of the first working set, the training nodes performing training using the data items of a second working set that are in their shards; and while the training nodes are performing training on their respective subsets of shards of the second working set, the nodes randomly shuffling data items in the first working set to create a shuffled first working set.

Abstract: Systems and methods are configured to provide lifetime data valuations for a dataset that evolves across multiple machine learning training tasks by providing and updating path-dependent data valuations for data points in the dataset during each training task. A current machine learning training task may include splitting the dataset into multiple random mini-epochs and training the current machine learning model using a first random mini-epoch and an accuracy mini-epoch, which consists of high value data points from the path-dependent data valuations. The random and accuracy mini-epochs can be, during the training, iterated for a number of times during the training, while a second random mini-epoch is prefetch. During the training, the path-dependent data valuations can be updated based on data valuations during the current training and a similarity between the current machine learning model and prior trained machine learning models.

Abstract: Provided herein are methods for inhibiting ALPK1 kinase activity using a compound of Formula I, and compositions and methods for therapy, for example in treating Kawasaki Disease.

Abstract: Provided herein are compounds of Formula (I) or (II) and related compositions and methods for their use as inhibitors of alpha-kinase 1 (ALPK1) and in treating Kawasaki Disease.

Abstract: Systems and methods are configured to provide lifetime data valuations for a dataset that evolves across multiple machine learning training tasks by providing and updating path-dependent data valuations for data points in the dataset during each training task. A current machine learning training task may include splitting the dataset into multiple random mini-epochs and training the current machine learning model using a first random mini-epoch and an accuracy mini-epoch, which consists of high value data points from the path-dependent data valuations. The random and accuracy mini-epochs can be, during the training, iterated for a number of times during the training, while a second random mini-epoch is prefetch. During the training, the path-dependent data valuations can be updated based on data valuations during the current training and a similarity between the current machine learning model and prior trained machine learning models.

Abstract: The disclosure provides compositions and methods related to activating alpha-kinase 1 (ALPK1) for modulating an immune response and treating or preventing cancer, infection, inflammation and related diseases and disorders as well as potentiating an immune response to a target antigen. The disclosure also provides heterocyclic compounds of formula (I) as agonists of alpha protein kinase 1 (ALPK1) and their use in activating ALPK1, modulating an immune response and treating diseases such as cancer, wherein A1, A2, L1, L2, L3, Z1, Z2, W1, W2, R1, R2, R3, R4, R5, R6 and R7 are defined herein.

Abstract: Provided are compounds of Formula (I) or (II) and related compositions and methods for their use as inhibitors of alpha-kinase 1 (ALPK1).

Abstract: Provided are compounds of Formula I, compositions and methods for their use as inhibitors of alpha-kinase 1 (ALPK1).

Abstract: A load balancing method, device and system, where the method includes: determining, based on load statuses of respective probe channels on n paths between a source end and a destination end, a target path with a lightest load in the n paths, where a probe channel is used to transmit a probe packet that includes a bandwidth probe packet; sending a bandwidth probe packet to the destination end through a probe channel on the target path; receiving the bandwidth probe packet returned by the destination end; and sending, based on the bandwidth probe packet returned by the destination end, a to-be-transmitted data packet to the destination end through a data channel on the target path. Therefore, load balancing among a plurality of paths is implemented.

Abstract: In an embodiment, a source node assigns a first number to a probe data packet in a probe data flow in a sending sequence, where the first number is used to select a transmission path for the probe data packet. The source node sends the probe data packet in the probe data flow to a destination node at a first sending rate. Each time the source node receives a probe data packet backhauled by the destination node, the source node sends a service data packet in a service data flow, where the service data packet is assigned a second number corresponding to the probe data packet, and the second number is used to select a transmission path for the service data packet.

Abstract: Provided herein are CD47 antibodies and immunologically active fragments thereof that have low immunogenicity in humans and cause low or no level of red blood cell depletion or hemagglutination. Also provided are as well as pharmaceutical compositions containing such antibodies or antibody fragments, as well as methods of treatment using such antibodies, e.g., as single agents or in combination with other therapeutic agent (s).

Abstract: The disclosure provides compositions and methods related to activating alpha-kinase 1 (ALPK1) for modulating an immune response and treating or preventing cancer, infection, inflammation and related diseases and disorders as well as potentiating an immune response to a target antigen. The disclosure also provides heterocyclic compounds of formula (I) as agonists of alpha protein kinase 1 (ALPK1) and their use in activating ALPK1, modulating an immune response and treating diseases such as cancer, wherein A1, A2, L1, L2, L3, Z1, Z2, W1, W2, R1, R2, R3, R4, R5, R6 and R7 are defined herein.

Abstract: The disclosure provides compositions and methods related to activating alpha-kinase 1 (ALPK1) for modulating an immune response and treating or preventing cancer, infection, inflammation and related diseases and disorders as well as potentiating an immune response to a target antigen. The disclosure also provides heterocyclic compounds of formula (I) as agonists of alpha protein kinase 1 (ALPK1) and their use in activating ALPK1, modulating an immune response and treating diseases such as cancer, wherein A1, A2, L1, L2, L3, Z1, Z2, W1, W2, R1, R2, R3, R4, R5, R6 and R7 are defined herein.