79 results about "Model tuning" patented technology

The invention relates to the technical field of three-dimensional model adjustment, in particular to a three-dimensional model adjustment method and system and a medium. The method comprises the following steps: obtaining a multi-angle image of an original model, carrying out multi-view normalization on the multi-angle image, generating a normalized view image set, extracting feature points of the original model, carrying out parallax correction on the feature points, reconstructing a simulation three-dimensional model, collecting basic purpose data of the model, carrying out ideal demand mapping through the data, and carrying out ideal demand mapping. The method comprises the following steps: determining an ideal three-dimensional model structure, carrying out core region segmentation on a reconstruction model according to basic purpose data to obtain key region model slices, carrying out highlight region comparison with the ideal model structure, analyzing model differences, determining a structure adjustment amplitude interval according to a comparison result, and carrying out cyclic fine adjustment correction on the key region model slices to obtain a three-dimensional model. And the three-dimensional model is consistent with the ideal three-dimensional model in structure, so that the optimized three-dimensional model is generated. According to the invention, efficient and accurate three-dimensional model adjustment and optimization are realized.

The invention discloses an image classification and recognition system and method based on an AI training model, relates to the technical field of image classification and recognition, and solves the problem that an existing method lacks dynamic adaptability. Comprising a data monitoring module, a dynamic label correction module, a model distillation module, a model tuning module and a model version control module. The data monitoring module evaluates input data distribution offset in real time by adopting a maximum mean value difference and KL divergence algorithm, and triggers model updating; the dynamic label correction module realizes automatic label correction through clustering active learning and semantic segmentation; the model distillation module uses adaptive incremental learning and a knowledge distillation mechanism to reduce computing resource consumption; the model adjusting and optimizing module adopts an elastic weight integration method to maintain the stability of the model; performance fluctuation is prevented through the model version control module; according to the invention, the flexibility and response speed of electric power facility inspection are greatly improved, and the defect detection precision and the long-term adaptability of the model are remarkably improved.

The invention discloses an ice condition prediction method based on multi-feature fusion and physical constraint, and belongs to the technical field of hydrological forecasting. Firstly, various types of data of a target area are collected and preprocessed to serve as a data set, features of the various types of data are extracted, feature fusion is conducted on obtained image features, time sequence features and environment features through a multi-head attention mechanism, and a fusion feature vector is generated. Secondly, adopting a physical information neural network model, taking the fusion feature vector as input, taking ice thickness and ice stress as output layers, carrying out constraint by using a composite loss function, carrying out model optimization by using a verification set, carrying out processing through a full connection layer in the network, and carrying out end-to-end training and regularization of a prediction model; and finally, evaluating the final prediction model obtained by training through the test set, and verifying the effectiveness and generalization ability of the final prediction model. The method combines multi-source and multi-mode observation data, can effectively capture complex relations between ice surface changes and various factors, can improve prediction precision, stability and reliability, and can improve model training efficiency and generalization ability.

A combined hyperparameter and proxy model tuning method is described. The method involves multiple search iterations. In each search iteration, candidate hyperparameters are considered. An initial (‘seed’) hyperparameter is determined, and used to train one or more first proxy models on a target dataset. From the first proxy model(s), one or more first synthetic datasets are sampled. A first evaluation model is fitted to each first synthetic dataset, for each candidate hyperparameter, enabling each candidate hyperparameter to be scored. Based on the respective scores assigned to the candidate hyperparameters, a candidate hyperparameter is selected and used to train one or more second proxy models on the target dataset

The invention discloses a method for randomly predicting and assigning parameters of an excavated soil body of a beach-overtopping foundation pit. The method comprises the following steps of: collecting geological survey data and construction information of an excavated area of the beach-overtopping foundation pit, screening redundant parts of characteristic parameters, retaining variables which have influences on mechanical parameters of the soil body, and constructing a complete original database; collected data is divided into a training set and a verification set, the training set is used for preliminary training of the model, and the verification set is used for optimization and verification of the model. According to the method, the nonlinear relation of the complex variables is captured through the random forest model, dynamic monitoring and real-time optimization are combined, the soil parameter prediction precision is remarkably improved, and the construction uncertainty risk is reduced; the dynamic monitoring function quickly responds to soil body changes in construction, the prediction result is optimized, and the construction safety is improved; the random field assignment method effectively reflects the spatial variability of soil parameters, supports multi-scene risk assessment, and provides scientific basis and reliable guarantee for engineering design and construction.

A combined hyperparameter and proxy model tuning method is described. The method involves multiple search iterations. In each search iteration, candidate hyperparameters are considered. An initial ('seed') hyperparameter is determined, and used to train one or more first proxy models on a target dataset. From the first proxy model(s), one or more first synthetic datasets are sampled. A first evaluation model is fitted to each first synthetic dataset, for each candidate hyperparameter, enabling each candidate hyperparameter to be scored. Based on the respective scores assigned to the candidate hyperparameters, a candidate hyperparameter is selected and used to train one or more second proxy models on the target dataset.

The invention discloses an intelligent task planning method and device based on environment perception, equipment and a medium, and the method comprises the steps: fusing abstract environment information into a large language model, and carrying out the fine adjustment of the large language model, so as to enhance the environment perception capability of the large language model; inputting a task target and an environment state table into the fine-tuned large language model, and generating a multi-step action sequence with the highest probability through a step-by-step beam search strategy; according to the generated action sequence, the environment data state table is updated by simulating interaction between the robot agent and other objects, and therefore interaction tracks corresponding to the steps are generated. According to the method, the model is adjusted into an environment perception large language model fine tuning structure, so that higher accuracy and practicability are brought to intelligent task planning; the large language model can explore a wider search space by utilizing a step-by-step beam search strategy, so that more comprehensive search is beneficial to generating a more reasonable and effective plan.

The present disclosure is directed toward systems, methods, and non-transitory computer readable media that provide a contextual content generation system that trains and implements a unique machine learning architecture to generate context-specific digital content items based on a digital guideline document. In particular, the disclosed systems select a content generation method from among prompt engineering and / or updating one or more machine learning models to generate digital content. For example, the disclosed systems utilize machine learning models to extract key elements from a digital guideline document comprising context-specific guidelines for digital content. Further, the disclosed systems generate an augmented prompt comprising indications of key elements from the digital guideline document. In addition, the disclosed systems select a content generation method from among prompt engineering and / or updating machine learning models to generate the digital content item which incorporates digital content corresponding to the context-specific guidelines based on the augmented prompt.

A method for model tuning and a system therefor are provided. The method according to some embodiments may include obtaining a deep learning model into which an adapter is inserted, wherein the adapter is a module inserted to adjust the output of the deep learning model using a plurality of weight matrices, and the plurality of weight matrices include a diagonal weight matrix, calculating loss by performing a specific task on the deep learning model and updating the plurality of weight matrices of the adapter based on the loss.

The invention provides a large model tuning method and system based on multi-modal information and AI, and relates to the technical field of artificial intelligence, and the method comprises the steps: extracting multi-modal parameter features and interaction features, and dynamically adjusting a contribution coefficient to achieve feature migration, and obtaining a cross-modal fusion vector; aligning the modal features by adopting an adversarial mechanism to obtain domain invariant features; constructing a prompt chain based on the modal confidence, optimizing the prompt chain by using AI reinforcement learning, and iterating to converge to obtain an optimization result; and finally updating the large model parameters. According to the method, effective fusion of multi-modal information is realized, and the tuning effect and generalization ability of the man-machine interaction large model are improved.

Auto-tuning covariances associated with a set of noise models for a variety of sensor modalities and / or perception components such that the covariances are leveled respective to one another may include whitening the covariances and / or error models and determining scalars to apply to the covariances. Determining these scalars may comprise using the residuals that result from generating the set of noise model (e.g., such as may be determined as part of least squares estimation) along with the hat matrix of the process model to determine the scalars. The covariances may iteratively be updated until the scalar adjustments converge or until another end condition is met.

Embodiments of the present disclosure provide techniques for adaptive tuning of motion stabilization models using artificial intelligence. A motion stabilization model for use with a device may be identified. Context data associated with a device may be identified. Metadata comprising one or more model parameters for the motion stabilization model may be identified. Model adjustment data may be generated based on the context data and the metadata by applying the context data and the metadata to a machine learning tuning model. The model adjustment data may be applied to the motion stabilization model to tune the motion stabilization model. The motion stabilization model may be configured to facilitate motion stabilization to account for screen motion of the display of the device and eye motion of a user relative to each other in a vehicle.

Systems and methods for model tuning and inference can include leveraging quality signals for determining resources to be more closely utilized during training and inference. The systems and methods can include obtaining a training dataset that includes a plurality of content items associated with a plurality of resources. The systems and methods can determine a plurality of quality scores for the plurality of resources. The systems and methods may then tune the parameters of the generative model based on the plurality of content items and the plurality of quality scores. The plurality of quality scores may be leveraged to train a reward model to be utilized for model-generated output selection.

The invention relates to the technical field of power prediction of a power system, and discloses a sample adaptive iterative training method and system for existing model tuning, and the method comprises the steps: obtaining the net load and solar irradiance data of a power distribution network; determining a payload-irradiance response intensity indicator based on the response slope within the local window; determining reference values of a truncation mode and a response mode according to statistical distribution of the indexes, and calculating a non-truncation probability of the sample in a normal power generation state; calculating the marginal response intensity of the prediction model to the solar irradiance; constructing a critical factor for capturing an inverter action boundary and a sample training weight by using a non-truncation probability; constructing a photovoltaic bistable constraint item, and constraining marginal response intensity to approach corresponding modal reference values in different states; and finally, performing parameter updating on the model based on the weighted prediction error and the bistable constraint term. According to the method, the problem of model tuning failure caused by equipment control truncation is effectively solved, and the prediction performance of the model under different operation conditions is ensured.

The invention discloses a model tuning method and device, equipment and a storage medium, and relates to the technical field of data processing, and the method comprises the steps: obtaining an initial data set, an initial question and answer pair comprising an initial question and a corresponding initial answer; in order to extract data related to the initial questions, first data corresponding to all the initial questions are retrieved from a preset database, then a target question-answer pair is determined by adopting a target model according to the first data and the initial question-answer pair, and the target question-answer pair is used as a target question-answer pair. The target question is a corresponding initial question which is not answered by the target model. In order to improve the performance of model tuning, the target data set is generated according to the target question and answer pair and the initial data set, then the target data set is used for tuning the target model, and the effective data in the target data set for tuning the target model is improved, so that the tuning effect of the target model is improved.

The invention provides a model updating method, device and system, in the model updating method, computer equipment displays a first processing result obtained by processing system data of a network system according to a model, and a user can perform knowledge data input operation after seeing the first processing result. The computer device obtains target knowledge data obtained after format conversion is conducted on the knowledge data, then the computer device responds to the model updating instruction, a model is updated through the target knowledge data, an updated model is obtained, and finally system data of the network system is processed according to the updated model to obtain a second processing result. And displaying the second processing result. The user can directly participate in the model adjusting and optimizing process, and the model of the user can be updated in real time to meet changing business requirements and personalized expectation of the user.

The invention provides a computing power asynchronous parallel fine-tuning large model method and device based on an intelligent computing center, and the method comprises the steps: obtaining a to-be-processed task, and determining a training set, a test set, a base large model and an evaluation index; determining a hyper-parameter space; based on the hyper-parameter space, the training set and the computing power resource of the intelligent computing center, performing parallel fine tuning on the base large model; based on the test set, the evaluation index and the computing power resource of the intelligent computing center, performing parallel evaluation on each fine-tuned large model to obtain an evaluation result corresponding to each fine-tuned large model; and determining an optimal evaluation result, determining the fine-tuned large model corresponding to the optimal evaluation result as a target large model, and recording a hyper-parameter group corresponding to the target large model. Through powerful computing power resources of the intelligent computing center and in combination with an asynchronous parallel processing mode, large model fine tuning and evaluation tasks based on multiple hyper-parameter combinations can be processed at the same time, the model tuning time is remarkably shortened, and the model tuning efficiency is improved.

The invention relates to the field of dialogue systems, and discloses an active dialogue model fine tuning method based on knowledge editing, a terminal and a medium. According to the method, an active dialogue data set and a universal pre-training data set in a specific field are collected and preprocessed, and an active dialogue task in the data set is divided into three sub-tasks of fuzzy demand prediction, active questioning and user demand overview; performing instruction fine tuning and secondary pre-training on the pre-training language model added with the parameter efficient module by using the active dialogue data set and the pre-training data set; specific capability is extracted and enhanced on the parameter efficient module, so that model parameters after adjustment of the pre-training language model are obtained; and based on the adjusted model parameters, triggering an active questioning ability through context learning, and realizing dynamic mining of user intentions. According to the method, the active dialogue model is fine-tuned, so that the potential intention of the user is mined through active questioning of the model, the understanding of the user on the performability of the task is improved, and subsequent tasks are guided to proceed.

The invention discloses a quality authentication method combining model fine tuning and a knowledge graph, belongs to the technical field of natural language processing, and aims to solve the problem that an existing quality authentication method based on a pre-training language model is insufficient in model performance, the method comprises the following steps: firstly, constructing a GLM model, and performing parameter fine tuning on the GLM model by adopting a low-rank adaptation method; the computing resource demand in the model fine tuning process is reduced, and the model tuning efficiency is improved. Meanwhile, by adopting integration of the knowledge graph, on one hand, the understanding and generation ability of the model to long texts and sentence levels is enhanced, and on the other hand, knowledge in the professional field of quality authentication is effectively integrated into the learning process of the model, so that the model can learn general knowledge from large-scale general texts and can also learn general knowledge from the sentence level. And moreover, professional tasks can be better completed by utilizing the injected knowledge in the quality authentication field, and the application effect of the model in the quality authentication field is improved through optimization and implementation of the model.

To tune a machine-learning model for a deployment system, a machine-learning model training system generates multiple tuning steps for the machine-learning model and generates an accuracy loss sensitivity for each tuning step. Each of the tuning steps indicate a corresponding set of one or more parameters and hyperparameters that reduces the impact of the machine-learning model on the system performance of the deployment system. Based on the accuracy loss sensitivities, the machine-learning model training system selects a tuning step with the least impact on the accuracy of the machine-learning model and modifies the machine-learning model based on the selected tuning step. After also modifying the tuned machine-learning model based on a threshold accuracy, the machine-learning model training system provides the tuned machine-learning model to the deployment system.

A kind of H&E image ovarian cancer polyploid giant cell (PGCCs) detection method based on deep learning, including constructing specific data set, image annotation and data division, using OCDet model to carry out feature learning and optimization, and model training and evaluation.The method first establishes the H&E staining image data set containing PGCCs, then accurately labels the image and divides it into training, verification and test set.OCDet model takes CSPDarkNet as the core, combines ECA mechanism, focuses on the deep learning and re-encoding of pathological semantic features.Through the training data set, the model updates parameters through back propagation and gradient descent, optimizes to identify PGCCs features.The verification set is used for model tuning, and the test set is used for evaluating the performance of the model.The automatic detection technology of the application can help doctors improve the diagnosis efficiency and reduce the error, provide an important reference for clinical treatment and prognosis evaluation, and show significant clinical application value.

The invention discloses a mechanism data hybrid process system modeling method based on a genetic programming algorithm, and the method comprises the steps: carrying out the induction and construction of a normative mechanism symbol system from the perspective of energy transmission for the hybrid modeling demands of the complementation of the mechanism and data in the process industry, and carrying out the modeling of the mechanism data hybrid process system based on the genetic programming algorithm. Furthermore, through three steps of a model generation step based on genetic programming, a model setting step based on heuristic Powell and a model evaluation step based on graph complexity, a high-quality explicit model considering fitting accuracy and structure simplification is autonomously constructed, so that a good twinborn analysis basis is provided for a complex industrial system data monitoring scene in the process industry.

The embodiment of the invention relates to the technical field of artificial intelligence, and discloses an LLM dialogue system-based intention recognition model tuning method, device and equipment and a storage medium, and the method comprises the steps: obtaining first sample labeled data and sample unlabeled data input by a user in an LLM dialogue system, and obtaining an initial intention recognition model; the model is used for performing intention recognition on information input by the user to predict the real intention of the user; based on the initial intention recognition model, the sample unmarked data and the first sample marked data, second sample marked data and error sample data are obtained; generating third sample marked data based on the error sample data and the large language model; and performing optimization processing on the initial intention recognition model based on the second sample marked data and the third sample marked data to obtain a target intention recognition model. The large language model is utilized to expand the sample data, and a large amount of manual annotation of the sample data is not needed, so that the manual annotation cost is reduced, and the recognition accuracy is improved.

The invention discloses an intelligent processing method and system for multi-modal data of an oral broadcast video based on deep learning, and relates to the technical field of oral broadcast video data processing, and the method comprises the following steps: 1, collecting multi-modal oral broadcast data in real time, recognizing emotion types, and carrying out the data preprocessing; step 2, constructing an oral playing animation generation model, judging the synchronization degree of the mouth shape, the head action and the audio in the oral playing animation, and taking coordinated regulation measures; 3, judging the consistency of the mouth shape and the emotion intensity in the oral playing animation, and implementing an anomaly repair measure; 4, constructing a depth time sequence prediction model, evaluating the difference degree between the actual oral playing animation and the prediction result of the depth time sequence prediction model, and carrying out video optimization and model tuning; the problems that when mouth shape supplement and animation filling are carried out by AI, precise synchronization of mouth shapes, expressions, head actions and emotional fluctuations is difficult to achieve, so that visual abnormity is caused, and the reality sense and the professional degree of videos are affected are solved.

The invention discloses a construction unit price intelligent prediction method based on a standard bill of quantities. The method comprises the following steps: S1, investigating and surveying influence factors on construction unit prices of the bill of quantities; s2, querying target data in a database, and cleaning the data according to rules; s3, data exploration; s4, performing feature screening; s5, carrying out statistics on all types of data containing each discrete feature, and coding each discrete feature by using a One-Hot method; s6, dividing a training set and a test set by adopting a random sampling method; s7, converting original data through feature engineering; s8, adjusting and optimizing the model; and S9, for the project quantity list of the construction unit price to be predicted, inputting the corresponding influence factors into the trained random forest model, and predicting to obtain a predicted value of the cost. According to the method, the problems of poor training effect and huge data training caused by not-in-place data cleaning are solved; and predicting the construction cost of the list granularity through a machine learning algorithm.

The present invention relates to the technical field of automatic circuit generation, and is a method and system for automatic circuit generation based on FPGA, comprising: inputting circuit design requirements into the module generation model to obtain an initial circuit function module set, normalizing the initial module characteristic parameter set to obtain a standard module characteristic parameter set, inputting the standard module characteristic parameter set into a module splicing model to obtain a module splicing circuit, performing simulation test according to an HDL file to obtain a simulation test result, judging whether the simulation test result meets the simulation test standard, if not, performing model tuning, if yes, performing performance test according to the HDL file to obtain a performance test result, judging whether the performance test result meets the performance test standard, if not, performing model tuning on the module splicing model, if yes, constructing an FPGA configuration file according to the HDL file. The present invention can solve the problems of long circuit design cycle and low efficiency.

The invention discloses a dynamic weighted sampling federated learning algorithm for an MEC system, and relates to the technical field of edge computing, and the algorithm comprises the following steps: an Internet of Things device collects local data, carries out the preprocessing and feature extraction, employs a dynamic weighted sampling strategy to adjust the sampling, a terminal device trains a model through a federated learning framework, and carries out the feature extraction of the local data; parameters are uploaded to a central server to be weighted and aggregated into a global model, the number of iterations is dynamically adjusted, parameters of the global model are analyzed and displayed in real time, and training process data are recorded. According to the method, data distribution in federated learning is optimized through a dynamic weighted sampling strategy, the efficiency and accuracy of model training are improved, the flexibility of an algorithm is enhanced by dynamically adjusting the number of iterations, the method adapts to the requirements of different scenes, global model parameters are analyzed and displayed in real time, training process data are convenient to monitor, optimize and record, and the training efficiency is improved. And a basis is provided for subsequent model tuning and troubleshooting, and the performance and practicability of federated learning in the MEC system are integrally improved.

Tuning result records indicating tuned schedules for machine learning tuning tasks are stored at a data store accessed from a tuning service. A given schedule indicates at least an order in which sub-operations of a tuning task are to be executed. In response to determining that the data store does not include a result record whose tuning task meets a similarity criterion to a tuning task determined from a tuning request, performance predictions for a set of candidate schedules are obtained using a performance prediction model, without running code which executes the candidate schedules. In a response to the tuning request, executable code corresponding to a preferred schedule identified from the candidate schedules using the performance predictions is included.

The invention discloses a self-supervised Chinese character generation method introducing dynamic convolution to enhance font style extraction, which is suitable for the fields of computer vision, font design and calligraphy education digitization, and comprises the following steps: constructing a dynamic convolution enhanced self-supervised network model which introduces a dynamic convolution kernel into the font style extraction field, and the convolution kernel parameters are dynamically adjusted to adapt to the edge and structure characteristics of the Chinese characters of different styles, so that the accurate capture of the style information is realized. According to the method, font style information extraction is enhanced through dynamic convolution, and the problems that in a traditional font generation method, a fixed convolution kernel is depended, the kernel scale and weight need to be manually adjusted, and the fixed kernel and a style extraction target are obviously separated, so that the model adjusting and optimizing process is tedious, and the cross-style generalization ability is poor are solved; therefore, the feature extraction adaptability and accuracy of different styles of Chinese characters are remarkably improved.

A method of tuning a machine learning model associated with a lithography patterning process is disclosed. More particularly, a method of using a reinforcement learning tuning engine to accurately adjust a deep learning-trained lithography model and enable global view of a training dataset is disclosed. In the reinforcement learning process, the reward can be defined as a function of global metrics derived from model predictions. Given the reward, the accuracy agent can determine actions corresponding to delta variables of the model.