Abstract: The present application relates to a network interface card configuration method and apparatus, an electronic device, and a storage medium. The method includes: associating a target pod group with a target network interface card identifier during generation of the target pod group; obtaining a first required acceleration card resource corresponding to the target pod group in response to the fact that the target network interface card corresponding to the target network interface card identifier is the target virtual network interface card; and according to the first required acceleration card resource, configuring the target virtual network interface card for the target pod group, or mapping the target virtual network interface card to the target physical network interface card corresponding to the target virtual network interface card, and configuring the target physical network interface card corresponding to the target virtual network interface card for the target pod group.

Abstract: A computer-implemented method of Named Entity Recognition (NER) includes receiving an input, identifying a plurality of candidate entities corresponding to the input, assigning word embeddings to the input at an embedding layer, capturing sequential context of the word embeddings in an encoding layer to obtain encoded word embeddings, constructing an entity relation graph using global coreference relations and local dependency relations to obtain a coreference graph and a dependency graph, fusing the encoded word embeddings, coreference graph, and dependency graph, via a graphical neural network (GNN), to obtain updated word embeddings, and decoding the updated word embeddings via a decoding layer to obtain enriched entity predictions.

Abstract: The present invention relates to a steroid compound, a use thereof and a preparation method therefor. It is expected that such compound can effectively treat mental and neurological diseases, and has good active efficacy, pharmacokinetic (PK) performance, oral bioavailability, stability, safety, clearance rate, and/or metabolic performance and the like.

Abstract: The present invention discloses a multi-modal design information unified expression and reasoning method based on a large language model. The present invention utilizes the large language model to extract corresponding information of function-behavior-structure from a design scheme, so as to construct a graph network of the function-behavior-structure, thereby achieving a relatively accurate unified expression of multi-modal design information. The present invention obtains differences in three dimensions of the function-behavior-structure based on comparison of graph networks at different moments.

Abstract: A loading/unloading system for a quay type full-automatic container terminal includes a plurality of shore cranes, an operation lane area between two rails of the shore cranes, an operation area from a rear side of a landside rail of the shore cranes to a yard, an automatic container yard area, an operation lane area of the yard, and facilities behind the yard. The shore cranes are disposed in parallel at a front edge of a container terminal to autonomously complete shipping and unshipping operations of containers, and autonomously complete loading/unloading processes of artificial intelligence transportation robots through information interaction with an artificial intelligence transportation robot system. The operation lane area between two rails of the shore cranes includes: a lambdoidal reverse operation area of inner container trucks, a ship lofting operation area, and a loading/unloading operation area of the inner container trucks, which are physically isolated by fences.

Abstract: Disclosed are a network card communication method and apparatus for an AI training platform, a device, and a medium. The method includes: building a switch network topology on the basis of a spine-leaf network, and configuring a preset number of virtual local area networks for each leaf switch in the switch network topology; virtualizing a physical network card to obtain virtual network cards, allocating the virtual network cards to corresponding job-containers according to a preset allocation rule, and allocating, to each virtual network card in the job-containers, different sub-networks corresponding to the virtual local area networks; and adding a corresponding sub-network communication policy routing rule to a pod where each job-container is located, whereby a virtual network card in the job-container sends training data to the remaining virtual network cards on a basis of the sub-network communication policy routing rule.

Abstract: The disclosure provides a processing method for a priority notification of an incoming call and a mobile device. The processing method for a priority notification of an incoming call is applied to a mobile device. The processing method includes: receiving an incoming call signal and generating a notification signal when a plurality of trigger conditions is met. The trigger conditions include: determining, according to the incoming call signal, that a repeated call is within a first preset time or a time interval between repeated calls is less than a second preset time, that a caller number corresponding to the incoming call signal is in a priority contact list, and that the mobile device is in a preset use situation. Then, a mandatory reminder mode is activated according to the notification signal to notify a user.

Abstract: The present invention relates to a steroid compound, a use thereof and a preparation method therefor. It is expected that such compound can effectively treat mental and neurological diseases, and has good active efficacy, pharmacokinetic (PK) performance, oral bioavailability, stability, safety, clearance rate, and/or metabolic performance and the like.

Abstract: A working method and device for a deep learning training task. GPUs are allocated to multiple deep learning training tasks according to the remaining resources of the GPUs in a single server node or multiple server nodes to achieve the effect of considering multiple deep learning training tasks while ensuring the utilization rate of the GPUs. The method comprises: obtaining a deep learning training task parameter input by a user; determining the type of the deep learning training task from the task parameter, the type of the deep learning training task type comprising: single model and multi-model; selecting GPUs by different policies according to different deep learning training task types; and selecting, according to the position of the GPU, a CPU having a shortest communication distance from the GPU for working.

Abstract: A method of resource allocation includes: when it is detected that a target training task in a target development environment is started, triggering loading of a client plugin library. Therefore, the client plugin library may redirect a loading process of a target deep learning framework for the target training task, to hijack a startup process of the deep learning framework, and a target graphics processing unit request is generated during this process to request allocation of graphics processing unit resources for the target training task. Compared to the prior art, this embodiment of the present disclosure starts from a perspective of the deep learning framework, analyzes a loading logic of the deep learning framework when the training task is started, and achieves dynamic graphics processing unit sharing through hijacking the framework.

Abstract: The present disclosure provides a configuration method for virtual network interface card resource, including: integrating a plurality of virtual network interface cards in a node where a container group is located, so as to obtain a virtual network interface card set including a plurality of virtual network interface card groups; assigning a target virtual network interface card group for the container group from the virtual network interface card set; analyzing the target virtual network interface card group to obtain address information of target virtual network interface cards in the target virtual network interface card group; and configuring a virtual network interface card resource for the container group according to the address information.

Abstract: The disclosure provides an intelligent horizontal transportation system for automatic loading or unloading at a container terminal. The system includes a plurality of autonomous transport robots (ATRs) and an ATR control system. The plurality of ATRs are configured to perform horizontal transportation tasks. The ATR control system is in real-time communication with the plurality of ATRs to manage and control the plurality of ATRs; the ATR control system is in real-time communication with a terminal management system, an automated yard crane, and an automated quay crane, so as to coordinate task scheduling between the automated yard cranes, automated quay cranes, and the plurality of ATRs. The ATR control system includes a task scheduling module, a dynamic path planning module, a standardized control interface module, a traffic management module, a lock station management module, a vehicle sequencing module, a charging scheduling module, a parking management module, and a remote driving module.

Abstract: A method of resource allocation includes: when it is detected that a target training task in a target development environment is started, triggering loading of a client plugin library. Therefore, the client plugin library may redirect a loading process of a target deep learning framework for the target training task, to hijack a startup process of the deep learning framework, and a target graphics processing unit request is generated during this process to request allocation of graphics processing unit resources for the target training task. Compared to the prior art, this embodiment of the present disclosure starts from a perspective of the deep learning framework, analyzes a loading logic of the deep learning framework when the training task is started, and achieves dynamic graphics processing unit sharing through hijacking the framework.

Abstract: Disclosed are a network card communication method and apparatus for an AI training platform, a device, and a medium. The method includes: building a switch network topology on the basis of a spine-leaf network, and configuring a preset number of virtual local area networks for each leaf switch in the switch network topology; virtualizing a physical network card to obtain virtual network cards, allocating the virtual network cards to corresponding job-containers according to a preset allocation rule, and allocating, to each virtual network card in the job-containers, different sub-networks corresponding to the virtual local area networks; and adding a corresponding sub-network communication policy routing rule to a pod where each job-container is located, whereby a virtual network card in the job-container sends training data to the remaining virtual network cards on a basis of the sub-network communication policy routing rule.

Abstract: The present disclosure provides a configuration method for virtual network interface card resource, including: integrating a plurality of virtual network interface cards in a node where a container group is located, so as to obtain a virtual network interface card set including a plurality of virtual network interface card groups; assigning a target virtual network interface card group for the container group from the virtual network interface card set; analyzing the target virtual network interface card group to obtain address information of target virtual network interface cards in the target virtual network interface card group; and configuring a virtual network interface card resource for the container group according to the address information.

Abstract: A mobile video and audio box control system includes a television (TV) stick, a video splitter, a converter of High Definition Multimedia Interface (HDMI) to DisplayPort (DP), an interface of HDMI to red, green, and blue (RGB), a sub screen, a Micro Control Unit (MCU) circuit, a power management circuit, and a power circuit. The TV stick transmits the video and audio signals to the converter of HDMI to DP and the interface of HDMI to RGB by the video splitter, and the interface of HDMI to RGB transmits the video and audio signals to the sub screen; the MCU circuit is configured to control the interface of HDMI to RGB; the power management circuit is configured to supply power. The system is integrated and stable, and users can watch video and audio contents anytime and anywhere. The system has easy operation, good practicability, and strong battery life.

Abstract: A dry adhesive including a non-woven fabric of nanofibers comprising a polymeric material. The nanofibers include an average diameter within a range of from 50 nm to 10000 nm, and the non-woven fabric of nanofibers has a thickness within a range of from 0.1 ?m to 100 ?m. The dry adhesive has a shear adhesion strength that is higher than the normal adhesion strength. Additionally disclosed is a dry adhesive fiber mat including substantially aligned fibers extending lengthwise from a first region to a second region, each of the fibers having a diameter that is less than or equal to 10 microns. The dry adhesive fiber mat is selectively repositionable to, from, and between a mechanically interlocked state wherein the fibers at the first region are intimately positioned within void spaces between the fibers at the second region and a separated state not intimately positioned as such.

Abstract: A computer-implemented method of Named Entity Recognition (NER) includes receiving an input, identifying a plurality of candidate entities corresponding to the input, assigning word embeddings to the input at an embedding layer, capturing sequential context of the word embeddings in an encoding layer to obtain encoded word embeddings, constructing an entity relation graph using global coreference relations and local dependency relations to obtain a coreference graph and a dependency graph, fusing the encoded word embeddings, coreference graph, and dependency graph, via a graphical neural network (GNN), to obtain updated word embeddings, and decoding the updated word embeddings via a decoding layer to obtain enriched entity predictions.

Abstract: A working method and device for a deep learning training task. GPUs are allocated to multiple deep learning training tasks according to the remaining resources of the GPUs in a single server node or multiple server nodes to achieve the effect of considering multiple deep learning training tasks while ensuring the utilization rate of the GPUs. The method comprises : obtaining a deep learning training task parameter input by a user, determining the type of the deep learning training task from the task parameter, the type of the deep learning training task type comprising : single model and multi-model; selecting GPUs by different policies according to different deep learning training task types; and selecting, according to the position of the GPU, a CPU having a shortest communication distance from the GPU for working.

Abstract: The invention discloses a multi-step prediction method and system of future wind speed based on automatic reservoir neural network, realizes accurate and fast multi-step prediction of future information, maintains high robustness to noise and system time-varying, and avoids over-fitting problems. The technical scheme is: for short-term high-dimensional wind speed data, based on the delay embedding theory, the observed high-dimensional dynamics is used as the reservoir by using space-time information transformation, and the high-dimensional wind speed data is mapped to the future information of the target variable. The automatic reservoir neural network realizes the multi-step prediction of the target variable by solving a pair of conjugate space-time information interaction equations.