Abstract: Provided are a method for processing model parameters, and an apparatus. The method comprises: a model parameter set to be sharded is obtained, wherein the model parameter set comprises a multi-dimensional array corresponding to a feature embedding; attribute information for a storage system used for storing the model parameter set to be sharded is obtained, wherein the storage system used for storing the model parameter set to be sharded differs from a system on which a model corresponding to the model parameter set to be sharded is located when operating; the model parameter set to be sharded is stored in the storage system according to the attribute information.

Abstract: A memory device includes a memory array comprising a plurality of memory elements and a memory controller coupled to the memory array. The memory controller when in operation receives an indication of a defect in the memory array determines a first location of the defect when the defect is affecting only one memory element of the plurality of memory elements, determines a second location of the defect when the defect is affecting two or more memory elements of the plurality of memory elements, and performs a blown operation on a defective memory element at the second location when the defect is affecting two or more memory elements of the plurality of memory elements.

Abstract: A cavity creation tool by crushing with multi-stage controllable water jet, it is used in natural gas hydrate development and mainly consists of an inner tube upper joint, an inner tube lower joint, an intermediate sleeve, an inner structure consisting of a coaxial throttle push rod, an outer layer sleeve, an outer layer structure consisting of a supporting ring, a jet head mounted to the intermediate sleeve and threading the outer layer sleeve, and a jet crushing structure consisting of a single-stage telescopic jet head and a two-stage telescopic jet head.

Abstract: A network connection method and device for a training participant of a joint training model are provided, the training participant operates in a master-worker mode. The method includes: acquiring communication state information of a worker, the communication state information indicating a communication connection phase that the worker is in; acquiring communication state information of a target worker as target communication state information, where the target worker includes a peer node corresponding to the worker, and the peer node belongs to a different training participant of the joint training model; and resetting, in response to determining that the target communication state information does not match the communication state information of the worker, a communication connection phase that the worker is in.

Abstract: Provided are a method and apparatus for evaluating a joint training model. A specific implementation of the method for evaluating a joint training model comprises: receiving a model evaluation data request sent by a target device, wherein the target device comprises a participant of a joint training model; acquiring a sample set matching the model evaluation data request, wherein the matching sample set is labeled data associated with the joint training model; and generating model evaluation data of the joint training model according to the matching sample set. By means of the implementation, an effect index of a joint training model can be shared on the premise of not exposing original sample data. Accordingly, a timely and effective data reference basis is provided for the optimization and improvement of the joint training model.

Abstract: Provided are a method and an apparatus for generating prediction information, an electronic device, and a computer readable medium. The method includes: generating, based on first user characteristic information of a target user, anonymous user information of the target user (201); sending the anonymous user information to a second processing end to enable the second processing end to generate prediction information based on the anonymous user information and second user characteristic information (202) of the target user. Data interaction and sharing are realized while ensuring data privacy, thereby improving accuracy of the prediction information.

Abstract: A full-digital control rod position measurement device and a method thereof. The full-digital rod position measurement device transforms the core process of rod position measurement that is normally processed by an analog circuit or analog-to-digital mixed circuit into a digital processing. The full-digital rod position measurement device comprises an excitation power supply, an integrated interface board, and a universal signal processor. The universal signal processor processes output signals of detectors according to a preset numerical processing algorithm and outputs Gray code rod position signals. The full-digital rod position measurement device and method disclosed by the present disclosure may effectively reduce the complexity of the primary excitation circuit and the secondary measurement signal processing circuit of the detectors, and improve the operation reliability and measurement accuracy of the rod position processing equipment.

Abstract: Disclosed is an oil extraction and gas production method capable of in-situ sand control and removal by downhole hydraulic lift achieved by downhole oil extraction and gas production system and ground oil extraction and gas production system. The downhole systems mainly comprises a double-layer tube, a double-layer tube reducing joint, a double-layer tube packer, a hydrodynamic turbine motor, a sludge screw pump, a soil-sand separator and a negative pressure absorber; the ground system comprises a power fluid pressurizing module and a mix fluid treatment module. The present application lowers the difficulty of pumping and lifting downhole formation fluid; achieves downhole and in-situ sand control and sand discharge, alleviates the blockage and erosion of sand particles on equipments and reduces energy consumption; decreases the production cost and improves the operation efficiency, therefore is suitable for oil extraction and gas production in high sand content wells.

Abstract: A trough plate for constructing a locked polymer anti-seepage wall includes a plate body, a guide tube, a grouting pipe and an anti-blocking head. A construction method of the locked polymer anti-seepage wall includes the steps of positioning an Nth trough plate and then pressing the Nth trough plate into ground, wherein N is a natural number larger than and equal to 1; engaging an (N+1)th trough plate with the Nth trough plate, and then pressing the (N+1)th trough plate into the ground; connecting a grouting pipe of the Nth trough plate with a grouting machine, pulling out the Nth trough plate, and simultaneously grouting through the grouting pipe of the Nth trough plate by the grouting machine; and repeating the steps (B) and (C) till the locked polymer anti-seepage wall is completed, wherein the steps (B) and (C) are repeated every time, N is automatically increased by 1.

Abstract: A system is disclosed for network management automation using network intent or adaptive monitoring automation. Network intent (NI) represents a network design and baseline configuration for that network or network devices with an ability to diagnose deviation from the baseline configuration. The NI can be automated to update and replicate the diagnosis. The monitoring of the network can be adapted to capture network problems in advance with adaptive monitoring automation.

Abstract: The disclosure discloses a solid fluidization tubular separator for marine natural gas hydrate, which includes a first separator and a second separator, wherein the first separator includes the first separation sleeve, the power liquid pipe, the swirl baffle, the recovery mechanism and the sand discharge mechanism. After the hydrate is sucked into the first separation sleeve to generate a circumferential velocity, so that the mud and sand with high density are separated to the pipe wall of the first separation sleeve, and the mud and sand separated to the pipe wall are settled down from the gap between the swirl baffle and the pipe wall along the pipe wall. The hydrate swirl flows into the recovery mechanism, and then leaves the first separation sleeve and enters the second separator, so as to realize the separation of mud and sand and natural gas hydrate.

Abstract: A trenchless integrative repair method includes steps of: S1, pre-casting steel rings according to an internal diameter of a defect pipeline, and forming grouting holes in the steel rings; S2, loading the steel rings to a defect position inside the pipeline, and fitting the steel rings to an internal wall of the pipeline to form reinforcing rings; S3, selecting a number of the grouting holes according to subsidence distribution; drilling a hole in the pipeline to reach a subsidence position and inserting a grouting pipe; steadily grouting the subsidence position with a polymer grouting system, and observing a lift condition of the pipeline; removing the grouting pipe after the pipeline is lifted to a predetermined level; and S4, pulling a lining pipe into the pipeline, and fitting the lining pipe to the steel rings and pipeline walls adjacent to the steel rings; curing the lining pipe.

Abstract: A data driver is provided, which includes a control unit, a data distribution unit, a first driving unit and a second driving unit. The data distribution unit is respectively coupled with the input terminal of the first driving unit and the input terminal of the second driving unit, and is configured to provide a first data distribution signal to the input terminal of the first driving unit and a second data distribution signal to the input terminal of the second driving unit; The first driving unit is configured to generate a first data signal according to a first data distribution signal, and the second driving unit is configured to generate a second data signal according to a second data distribution signal, and voltage polarities of the first data signal and the second data signal are different.

Abstract: The present disclosure discloses a magnetic flexible tactile sensor structure based on a folding magnetization method, which comprises a flexible body containing a permanent magnetic material; the flexible body has a negative Poisson's ratio structure, and its set area has undergone folding magnetization treatment. The present disclosure also discloses a sensor composed of the above-mentioned sensing structure. The sensor provided by the present disclosure can be applied in sealed and wireless scenarios. The present disclosure can detect the size and position of force. The flexible sensor has broad prospects in the application of the touch skin of robots. The function between the magnetic field-based sensing magnet and the Hall element is contactless. In some cases where it is difficult to establish isolation of the connection lines, it can also be used as an unfettered tactile sensor.

Abstract: A memory device includes a memory array comprising a plurality of memory elements and a memory controller coupled to the memory array. The memory controller when in operation receives an indication of a defect in the memory array determines a first location of the defect when the defect is affecting only one memory element of the plurality of memory elements, determines a second location of the defect when the defect is affecting two or more memory elements of the plurality of memory elements, and performs a blown operation on a defective memory element at the second location when the defect is affecting two or more memory elements of the plurality of memory elements.

Abstract: A method includes a first device and sends a network scanning request to a second device over a short-range wireless communication connection between the first device and the second device; the first device receives a scanning result that is of at least one first wireless signal and that is sent by the second device; and the first device sends a wireless connection request to at least one third device based on a wireless signal scanning result to establish a wireless connection to the at least one third device, where the wireless signal scanning result includes the scanning result of the at least one first wireless signal.

Abstract: Disclosed herein, are recombinant polypeptides comprising one or more homologous amino acid repeats fused with an IgG binding domain The recombinant polypeptides can be bound to a therapeutic antibody and used a delivery vehicle to increase the retention time and reduce systemic-related side effects of the therapeutic antibodies.

Abstract: The disclosure provides a thermal bending-free photocuring curved screen protective paster which comprises a first protective film, an OCA glue layer, a main functional layer, a PET film layer, a surface functional layer and a second protective film sequentially arranged from bottom to top. The protective paster of the disclosure can be well attached to a curved screen and the like, and avoids the phenomena of warping, forming failure, springback and the like occurring after the traditional protective film is attached. The protective paster of the disclosure has good wear resistance, good anti-scraping effect, good transparency, good temperature resistance and stable structure, which can be prepared into a film with any angle, can be firmly attached to a curved screen with a large radian, and can effectively protect the curved screen.

Abstract: A multi-device wireless connection method and a device. The method includes obtaining device information corresponding to a first account, selecting a second device according to a user-triggered selection instruction and the device information corresponding to the first account, where the second device is a device to which the first account is logged in, sending a pairing request to the server, where the pairing request comprises identification information of the second device, receiving pairing information sent by the server, where the pairing information is used for pairing between the first device and the second device, where the pairing information comprises a first random number which is generated by the server, or by the second device and sent to the server, and performing, by the first device, pairing with the second device according to the pairing information, and implementing a wireless connection between the first device and the second device.

Abstract: A system is disclosed for generating a graphical user interface (GUI) for network management automation. The GUI is a data view that provides functionality for monitoring and controlling devices in a network using a dynamic golden baseline.