393 results about "Model transformation" patented technology

The embodiment of the invention provides a multi-source heterogeneous data integration method and device fusing a large model conversion operator. The method comprises the steps that heterogeneous data are collected from a multi-source heterogeneous data source, preprocessing operation of data cleaning is carried out, and preprocessed data features are obtained; inputting the preprocessed data features and the target format into a large language model, and generating a conversion operator including data structure analysis, field mapping and type adaptation; and based on the conversion operator, performing distributed parallel data conversion and integration in a distributed computing framework, and integrating and verifying the integrated data. According to the scheme, by introducing the intelligent reasoning ability of a large language model, the efficiency optimization of distributed calculation and the quality verification mechanism of the whole process, the core problems of the traditional data integration technology in the aspects of rule stiffness, high manual dependency and quality control deficiency are systematically solved.

The invention discloses a three-dimensional automatic modeling method for a high-precision transformer substation. The method comprises the following steps: collecting laser point cloud data and multi-view pictures of the transformer substation; dense point cloud data registration, multi-view image camera parameter reconstruction, multi-view image projection and adaptive point cloud density control are carried out; approximating the point cloud data by using three-dimensional Gaussian distribution functions, summing all the three-dimensional Gaussian distribution functions to obtain a Gaussian distribution geometric field represented by the three-dimensional Gaussian distribution point cloud, and reconstructing and rendering a three-dimensional grid to obtain a three-dimensional grid model; setting constraint conditions of equipment on the basis of the three-dimensional grid model, and generating a three-dimensional equipment model, equipment coordinates and equipment classification names by using a volume segmentation algorithm and a three-dimensional marking algorithm; and converting the equipment model into a GIM model file, generating a PMS file from the equipment coordinate and the equipment classification name, and synchronizing the PMS file to a GIM and PMS 3.0 system. According to the method, the problems of low efficiency and inconsistent quality in the traditional manual modeling process can be effectively solved.

Systems and methods for unsupervised multi-modal causal structure learning for root cause analysis. System logs of a cloud system can be transformed to time-series data using a log-tailored language model to obtain system log features of the cloud system. A metric causal graph and a log causal graph can be predicted from modality-specific representations and modality-invariant representations of extracted system metric features and system log features, respectively, using the deep neural network. The metric causal graph and log causal graph can be fused to obtain a fused causal graph. Root causes of system failure can be flagged for system maintenance based on ranked entities obtained from the fused causal graph to obtain flagged root causes. System maintenance can be performed autonomously based on the flagged root causes from identified system entities to optimize the cloud system with an updated configuration.

Systems, apparatus, articles of manufacture, and methods are disclosed to utilize large language artificial intelligence models to convert computer code. An example apparatus includes instructions and processor circuitry to execute the instructions to at least: train a large language model based on a computer instructions repository that includes code of a first type; utilize the large language model to convert an input set of instructions of the first type into output code of a second type; cause execution of the output code; determine if the execution is successful; and when the execution is not successful, utilize the output code for fine-tuning training of the large language model with incorrect data.

In variants, the method can include generating mapping model training data, determining the mapping model, and predicting a program based on a transformer. The method can optionally include evaluating the mapping model, running analyses on the program, and / or utilizing the program and / or generated program analyses. The method functions to convert transformer models into programs that can be characterized and / or analyzed using program analysis techniques.

The invention relates to an NPU model optimization method and device based on TensorFlow and a storage medium. The method comprises the steps that a model analysis module reads and extracts the structure and parameters of a TensorFlow model; the model conversion module is used for converting the TensorFlow model into an intermediate representation supported by the NPU; through operator fusion, memory optimization and quantization, the model optimization module optimizes the converted TensorFlow model, and deploys the optimized TensorFlow model to the NPU execution module; and the performance monitoring module monitors the operation performance of the TensorFlow model on the NPU and outputs performance data. According to the method, an efficient model conversion tool is provided, the TensorFlow model is quickly converted into the format supported by the NPU, and the model conversion efficiency is improved; the hardware characteristics of low-precision calculation, parallel calculation and the like of the NPU are fully utilized, the calculation efficiency is improved, and the conversion time and the development cost are reduced by means of operator fusion, memory optimization, quantization and the like; manual operation of developers is reduced, the model deployment process is simplified, and the deployment difficulty is reduced; large-scale model deployment is supported, and actual application requirements are met.

The disclosure relates to utilizing a domain insight system for providing plain language descriptions and insights into complex data and / or sparsely populated domains using machine-learning models and large generative models. For instance, the domain insight system converts data outputs from machine-learning models in various output formats into clear, accurate, comprehensible, and straightforward results. The domain insight system achieves this by using one or more dynamic prompts that are tailored based on the data output types and report descriptors, thus improving the accuracy and efficiency of the large generative model. In particular, the domain insight system uses specialized prompts with carefully selected parameters and, in some cases, system-level meta-prompts, to generate accurate domain-based reports and explanations for a given dataset.

The present application provides a CAD model analysis method and apparatus, a device and a storage medium, which relates to the technical field of industrial manufacturing. The CAD model based on a BRep representation is transformed into a graph structure. The graph structure is input into an analytical model to obtain a modeling sequence of the CAD model. By transforming the CAD model with a BRep representation into the graph structure, the analytical model learns the modeling feature in the graph structure, and the modeling feature type and the feature parameter are obtained, so as to enable the prediction of the model building process of the CAD model and facilitate to subsequently make parametric re-editing of the CAD model

The invention discloses an electric power inspection robot detection and deployment method and system and computer equipment, and the method comprises the steps: obtaining high and low wire harness three-dimensional point cloud data of a to-be-detected road scene, inputting the data into a teacher model and a student model, extracting BEV features, employing a focus module to carry out the distillation processing of the three-dimensional features of the teacher-student model, and obtaining the three-dimensional point cloud data of the teacher-student model; creating foreground and background masks, and calculating the loss of foreground features; a focus relation distillation module is introduced, features of nine angular points of a foreground position are extracted, Gaussian similarity calculation is carried out, and the balance of the model in different types of foreground feature learning is improved; a global distillation module is applied to improve the overall performance, model pruning is performed after model training is completed, and the model performance is optimized; and the model after pruning optimization is converted into an ONNX format, deployment and acceleration are carried out on TensorRT, and the reasoning efficiency is further improved. The problems of low detection precision of the low-wire-harness radar and difficulty in actual deployment of the model are effectively solved.

An adapter to a base model of an artificial intelligence (Al) system is disclosed. The adapter includes a connector to connect the adapter to the base model such that during an operation of the Al system at least some portion of data transformed by the base model is propagated from the base model to the adapter and back from the adapter to the base model. The adapter includes a non-linear modifier to modify the data received from the base model non-linearly before returning the modified portion of the data back to the base model, and an Al trainer to tune the non-linear modifier of the adapter by propagating training data through the base model and the adapter and updating weights of the non-linear modifier of the adapter for given weights of the base model to optimize a loss function. Further, weight matrices for the base model and the adapter are jointly constructed by an additional module, which efficiently uses a pool of parameters to allocate to save memory requirement for adaptation of the Al system.

The invention relates to the technical field of artificial intelligence chips, and discloses a model performance estimation method and device and computer equipment, and the method comprises the steps: determining model configuration data and candidate distributed strategies of a target model; converting the original calculation graph corresponding to the single-card deployment state based on the model configuration data and the strategy configuration data of the candidate distributed strategies to obtain a distributed overhead calculation graph corresponding to the multi-card deployment state; and performing performance estimation on the candidate distributed strategy according to the basic overhead and the additional overhead in the distributed overhead calculation graph to obtain strategy performance data of the target model under the candidate distributed strategy, thereby realizing conversion of a single-card model into a multi-card model. And based on the multi-card model, simulation calculation of the multi-card interconnection mode is realized on the premise of limited hardware resources, so that the influence of a communication operator and a topological structure corresponding to the multi-card interconnection mode in the overall operation of the model is reflected, and the upper limit of the model performance can be accurately evaluated in a simulator verification stage before silicon is applied.

The invention discloses a rotor reliability constrained rolling bearing assembly parameter robust design method. The method comprises the following steps: constructing a dynamic model for an actual rotor-bearing system; constructing an uncertainty parameter vector and a design variable vector; a target function based on robustness and a constraint function based on reliability are constructed, so that an uncertainty optimization model is obtained; constructing an augmented input variable, and establishing a candidate orthogonal polynomial basis function set; on the basis, constructing and evaluating polynomial chaos-Kriging models for the target function and the constraint function respectively, and screening out an optimal polynomial chaos-Kriging model; calculating the expectation and the standard deviation of the target function and the failure probability of the constraint function under each design variable vector; and converting the uncertainty optimization model into an unconstrained single-target optimization model, randomly generating population individuals of a heuristic optimization algorithm in a feasible region of design variables, and iteratively searching an optimal solution of the unconstrained single-target optimization model as a rolling bearing assembly scheme.

The embodiment of the invention provides a model conversion method, a related device and a medium. According to the method, a model conversion mirror image is deployed to a node where a target chip is located, and an original model is converted into a target model corresponding to the target chip through the model conversion mirror image. And then the original model before conversion and the target model after conversion are verified through the original model verification mirror image and the target model verification mirror image respectively, a conversion deviation value capable of judging the model conversion effect is obtained according to the first verification result and the second verification result, and automatic verification of the target model after conversion is achieved. And based on the verification result (conversion deviation value), constructing a mirror image for the target model meeting the requirement, and storing the mirror image in a mirror image warehouse for subsequent calling. The embodiment of the invention aims to ensure the availability of the converted model. The embodiment of the invention can be applied to scenes of computer vision, natural language processing, voice recognition, automatic driving, edge calculation and the like.

The invention relates to the technical field of deep learning, in particular to a PyTorch framework-based compatibility expansion system, which comprises a cross-framework model converter, a heterogeneous hardware abstraction layer, a hybrid computational graph execution engine, an intelligent distributed trainer and a self-adaptive optimizer, the cross-frame model converter converts input model formats of different frames into PyTorch executable formats, the heterogeneous hardware abstraction layer supports rear ends of various hardware and automatically selects an optimal calculation path, and the hybrid calculation graph execution engine fuses dynamic graph flexibility and static subgraph optimization and supports dynamic control function execution flow. An intelligent distributed trainer automatically selects a parallel scheme and is compatible with multi-protocol communication, a self-adaptive optimizer performs dynamic optimization based on hardware characteristics and a PyTorch tensor, a cross-frame model converter supports model conversion of multiple deep learning frames, the cost of migration among different frames by a user is reduced, and the universality of the PyTorch frame is improved.

The invention discloses a source-network-load-storage coordinated optimization scheduling method and system based on multiple time scales, and the method comprises the steps: building a multi-time-scale optimization scheduling frame through defining scheduling stages of different time resolutions; on the basis of a multi-time-scale optimization scheduling framework, a mathematical model is established to describe an objective function and an optimization variable of each stage, and a multi-time-scale optimization scheduling model is established in combination with general constraint conditions; converting the model into a mixed integer second-order cone programming problem by utilizing a linear processing technology of convex relaxation and segmentation; solving the mixed integer second-order cone programming problem by adopting an optimization solver to obtain an optimization scheduling strategy adapted to different time scale demand changes; the method is not only suitable for a traditional power system, but also provides powerful support for development of a modern intelligent power grid, and ensures efficient, reliable and sustainable operation of the power system.

A processor-implemented method for converting a trained language model into a language model in an architecture of mixture of experts (MoE), and a computing device using the same is provided. The method for converting a trained language model into a language model in an architecture of mixture of experts using a computing device according to an embodiment of the disclosure may include dividing a plurality of layers included in a target language model and extracting a feed-forward network (FFN) included in each of the plurality of layers, generating an MoE block of the MoE language model, which corresponds to the feed-forward network, generating an input tensor, comparing output tensors between the feed-forward network and the MoE block for the input tensor to obtain a first loss, and updating a weight of the MoE block, based on the first loss.

Systems and methods are provided for adapting a pretrained language model to perform cybersecurity-specific named entity recognition and relation extraction. The method includes introducing a pretrained language model and a corpus of security text to a model adaptor, and generating a fine-tuned language model through unsupervised training utilizing the security text corpus. The method further includes combining a joint extraction model from a head for joint extraction with the fine-tuned language model to form an adapted joint extraction model that can perform entity and relation label prediction. The method further includes applying distant labels to security text in the corpus of security text to produce security text with distant labels, and performing Distant Supervision Training for joint extraction on the adapted joint extraction model using the security text to transform the adapted joint extraction model into a Security Language Model for name-entity recognition (NER) and relation extraction (RE).

A privacy protection heterogeneous federal learning method with Byzantine robustness includes training a local model and calculating a sketch, removing an abnormal local model, selecting a client with quick response capability, encrypting local update and submitting ciphertext, executing weighted aggregation and distributing an updated global model. The method has the beneficial effects that a high-dimensional local model is converted into a low-dimensional sketch by adopting locality sensitive hashing, and the quality of the local model is effectively evaluated on the premise of protecting data privacy. Then, the Byzantine clients are identified and removed based on the hierarchical clustering technology, aggregation weights are distributed to the remaining clients according to the local model quality, and the global model convergence speed is increased; in addition, by selecting the clients which are high in response speed and have representative data sets to participate in model training, the problem of outdated persons caused by system isomerism is solved under the condition that the global model accuracy is not affected.

The invention discloses a vehicle collision accident severity prediction method and device, equipment and a storage medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: collecting the multi-modal data of a vehicle collision accident, and carrying out the preprocessing of the multi-modal data; converting the pre-processed structured numerical data into a first feature map through a Grubrum angle field method; converting the preprocessed unstructured text data into a text semantic vector through a pre-training language model, and converting the text semantic vector into a second feature map; performing size alignment on the first feature map and the second feature map, and performing splicing on a channel dimension to obtain a multi-channel fusion image; and inputting the multi-channel fusion image into a first deep learning model, and outputting an accident severity prediction result. According to the method, the recognition sensitivity and the prediction recall rate of serious injury accidents are effectively improved, meanwhile, complex artificial feature engineering is avoided, and the generalization ability and the interpretability of the model are enhanced.

The invention discloses an application migration method and device, equipment, a readable storage medium and a program product, and relates to the technical field of artificial intelligence, and the application migration method comprises the steps that self-defined resources corresponding to reasoning applications are managed through a self-defined controller, so that automatic execution of a driving migration task is achieved. Specifically, on the basis of file conversion resources, an original model file is transformed; transforming an application mirror image based on the mirror image conversion resource; and migrating the reasoning application to a target resource pool corresponding to the target environment based on the application program migration resource. Namely, the migration process of the whole reasoning application is driven by a user-defined controller, and model transformation, mirror image adaptation, data migration and application deployment are automatically completed, so that efficient migration of the reasoning application among different environments is ensured. Therefore, the technical problem of cross-heterogeneous resource pool migration of the inference application can be solved, and the technical effects of getting rid of the dependence of cross-heterogeneous resource pool migration on manpower and guaranteeing the continuity and stability of services are achieved.

Described are a system, method, and computer program product for distributed execution of a machine-learning model on a server cluster. The method includes initiating retrieval of a machine-learning model from a data repository and converting the machine-learning model to an executable format. The method includes transmitting the converted machine-learning model to each node of the server cluster and executing the converted machine-learning model on each node. The method includes generating an initial performance metric based on execution of the converted machine-learning model on each node. The method includes transmitting the plurality of initial performance metrics from each node to an external processor and combining the plurality of initial performance metrics to produce a combined performance metric. The method includes modifying a model hyperparameter of the machine-learning model based on the combined performance metric and executing the modified machine-learning model in a computer system to evaluate real-time event data.

The invention discloses a heterogeneous object model conversion method and device based on a large language model, and the method comprises the steps: (1) an equipment model analyzer is responsible for loading and analyzing an object model file and a profile file of equipment, and obtaining the object model type of the equipment and the core function data of the equipment; (2) the object model mapper plays a crucial role in the object model conversion process, and a model mapping dictionary between different object models is obtained by learning the relation between different equipment object models and fine adjustment of related model mapping rules; (3) the object model converter performs object model macroscopic framework conversion based on the model mapping dictionary, converts source equipment object model core function data in the equipment model analyzer into target object model core function data, and fills an object model and a profile file of target equipment with the target object model core function data; in order to ensure the correctness of conversion, conversion detection is carried out on a converted target object model file and a profile file based on a large model. According to the method, the difference between object models can be overcome, and the universality of conversion is provided.

The invention discloses an end-to-end zero code AI model automatic training and deployment system and method, and belongs to the technical field of artificial intelligence and machine learning engineering, and the system comprises a user interaction layer which is used for providing a graphical user interface and receiving a model training instruction configured by a user; the business logic layer responds to the instruction, encapsulates the instruction into a structured work order, automatically executes data management, model training and model conversion tasks based on the work order, and generates a model deployment package capable of directly running on at least one target hardware platform; and the resource management layer is used for monitoring and scheduling bottom computing resources to support task execution of the business logic layer. According to the method, a Web-based graphical interface is provided, all operations are completed through clicking, dragging and pull-down selection, dependence on programming and command line operations is thoroughly eliminated, and complete AI workflow zero-code operation from data to deployment in a single system is achieved.

The invention discloses a product extension method based on a domain design mode, and the method comprises the steps: constructing a domain general framework according to the demands of a specific domain; based on the domain general framework, a platform independent model is defined, and the platform independent model comprises a domain general PIM and an enterprise special PIM; converting the platform-independent model into codes based on a code generator FCG of a domain general framework; and adjusting or expanding the field general PIM code and the enterprise special PIM code according to the new business requirement or function requirement to realize product expansion. And the consistency and maintainability of codes are ensured.

The invention relates to the technical field of formalized analysis and verification of a complex software system, and particularly provides a multi-modal cross-language multilayer verification formalized integration method. According to the method, a multi-dimensional analysis and verification process from codes to models is realized through a code or model self verifier layer, a code or model plug-in verifier layer, a multi-language or multi-model conversion verification layer and a code and model bidirectional cross verification layer. The multi-level verification framework covers code static analysis, model dynamic simulation and cross-language and cross-model conversion, various formalized tools are systematically integrated, verification results of different languages and models are deeply integrated while automatic and efficient verification is achieved, on one hand, system compatibility and consistency are kept, and on the other hand, the verification efficiency is improved; on the other hand, the sufficiency and reliability of system analysis and verification are remarkably improved.

The invention belongs to the technical field of electric power system dispatching optimization and electric power and electric quantity balance analysis, and relates to a time sequence random production simulation method considering a flexibility demand in an electric power market environment, which comprises the following steps: 1, constructing an electric power system time sequence-random production simulation framework considering the flexibility demand in the market environment; 2, the upper layer model obtains a start-stop scheduling scheme by taking the minimization of the total operation cost of the system as a target, calculates the time sequence flexibility supply capability of each type of unit, and provides an input basis for the lower layer; 3, constructing a multi-state flexibility supply model of the unit under different time scales in a lower layer model, carrying out multi-time scale system flexibility supply and demand probabilistic matching by adopting an equivalent electric quantity function method, and realizing flexibility index evaluation under a short time scale; 4, converting and solving the model; according to the invention, support is provided for coordinating the operation economy and flexibility of the system under security constraints.

The invention relates to the technical field of three-dimensional modeling, particularly provides a three-dimensional geometric model conversion method, a controller and a storage medium, and aims to solve the problem of how to improve the model conversion efficiency of a three-dimensional geometric model. In order to achieve the purpose, the method comprises the steps that an original model file, containing model data in an active format, of the three-dimensional geometric model is obtained, a target application scene suitable for the three-dimensional geometric model obtained after model conversion is obtained, data mapping and data optimization are conducted on the original model file on the basis of the target application scene, and the three-dimensional geometric model is obtained. And obtaining a target model file containing the model data in the target format. According to the method, the target application scene of the three-dimensional geometric model is obtained in advance, invalid data and incomplete data in the original model file can be filtered, the model data in multiple source formats are converted into the model data adaptive to the target application scene, the data mapping and data optimization speed is increased, and the data mapping efficiency is improved. And the model conversion efficiency and the scene application degree of the three-dimensional geometric model are improved.

The invention discloses a method for cooperative scheduling of an automatic assembly station and an AGV (Automatic Guided Vehicle). The method comprises the following steps: constructing a disjunction graph, and defining nodes, directed edges and undirected edges in the disjunction graph; an island type assembly workshop scheduling model is constructed, an objective function and constraint conditions are set, and the objective function is used for minimizing the maximum completion time; converting the island type assembly workshop scheduling model into a Markov decision process, setting a state space and an action space, and obtaining a reward function; an ORA-PPO algorithm network model is constructed, an environment state feature vector is input, the ORA-PPO algorithm network model is optimized according to the target function, parameters are updated, an optimal solution of the target function is obtained, and the environment state feature vector comprises elements in a state space and an action space. According to the invention, the problem of cooperative scheduling of multi-product and multi-process equipment and the AGV in the prior art is solved.