297 results about "Linear layer" patented technology

The invention relates to the technical field of model compression, and provides a large model compression method and device, a task processing method and equipment and a storage medium, and the large model compression method comprises the steps: carrying out the layer-by-layer quantification of a linear layer of a to-be-compressed initial large model, and obtaining a first large model; the initial large model is a pre-trained large language model constructed based on an expert hybrid architecture; performing route calibration on each expert sub-model in the first large model to obtain a second large model; in the reasoning process of the second large model, based on the task type of a to-be-executed target task, the importance of each expert sub-model in the task type is evaluated; and performing dynamic pruning on each expert sub-model based on importance so as to compress the second big model. Through a compression mode of combining static quantification and dynamic pruning, on the basis of ensuring the model performance, the memory and calculation overhead required by large model reasoning can be reduced, and efficient operation of the large model on light-weight equipment with limited video memory resources is facilitated.

The invention discloses a quantification method and reasoning method of a large language model and electronic equipment, and belongs to the technical field of large language models.The quantification method of the large language model comprises the steps that each linear layer to be quantized in the large language model is quantized; dividing a channel of the linear layer in the hidden layer dimension into a normal channel and an outlier channel; performing INT8 quantization on the first activation matrix corresponding to the normal channel in a word segmentation token dimension to obtain a second activation matrix, and performing INT4 quantization on the first weight matrix corresponding to the normal channel according to an output channel to obtain a second weight matrix; and determining an output result of the linear layer according to the second activation matrix, the second weight matrix, a third activation matrix corresponding to the outlier channel and a third weight matrix corresponding to the outlier channel.

The invention provides a model training method and device, equipment and a storage medium, and relates to the technical field of computers, in particular to the technical field of neural network models and model training. The specific implementation scheme is as follows: a calculation unit executes quantization matrix multiplication based on Hadamard pre-transformation on an activation tensor and a weight tensor of a target model stored in a memory so as to generate an output tensor of a linear layer based on a low-precision tensor with smaller data bit width; using the output tensor and a subsequent network layer of the target model to complete forward propagation so as to obtain a loss value; and according to the loss value, updating model parameters of the target model stored in a memory through a back propagation algorithm. By means of the technical scheme, on the premise that the model training precision is guaranteed, memory resource occupation and the calculation amount in the calculation process can be remarkably reduced, and the training cost is reduced.

The invention relates to a large model quantification method and system. The method comprises the following steps: performing greedy grid search on each linear layer of an off-line large model to obtain an optimal smoothing factor corresponding to each linear layer; performing reverse scaling on the weight and the activation value of the corresponding linear layer in the channel dimension by using the obtained optimal smoothing factor corresponding to each linear layer so as to keep the multiplication result between the scaled weight and the activation value unchanged, thereby obtaining the smoothed weight and the smoothed activation value; and performing quantization processing on the smoothed weight and activation value in the large model so as to convert into a numerical format of INT8 or FP8 or a numerical format lower than INT8 or FP8.

The invention discloses a cpu + gpu-based tensor parallel computing algorithm for a large model, which comprises the following steps of: S1, combining row parallelism and column parallelism for a linear layer in the model by adopting a mixed-dimension tensor segmentation mode, and dynamically adjusting a segmentation proportion according to a model structure and hardware resources; s2, a CPU and GPU cooperative computing mechanism is constructed, part of tasks which have large video memory requirements and are relatively simple in computation are allocated to the CPU, the GPU is responsible for computing intensive tasks, communication between CPU-GPU is optimized, and overlapping of CPU-GPU communication and GPU computation is achieved; and S3, implementing a dynamic resource allocation and load balancing strategy, monitoring load conditions of the CPU and the GPU in real time, dynamically adjusting task allocation according to calculation requirements of different layers of the model and use conditions of hardware resources, and adopting a self-adaptive batch processing size adjustment strategy. The invention obviously reduces the occupation of the video memory, reduces the communication cost and improves the utilization rate of computing resources.

The invention discloses a multi-wave-potential pulse radar target detection method and device based on deep learning and a medium, and the method comprises the steps: obtaining radar original echo data in different detection scenes, and forming a radar original echo data set; training the target detection network by using the radar original echo data set to obtain a target detection model; in an input preprocessing layer of the target detection network, a real part and an imaginary part of radar original echo data are separated, linear layer networks with different weight matrixes are used for carrying out feature mapping on the real part and the imaginary part of different wave position data, the features of the real part and the imaginary part are calculated, then the phase spectrum features are calculated, and a multi-dimensional feature map is formed; training a target detection network by using the echo data set; and obtaining echo data to be detected, and inputting the echo data to the target detection model for detection. According to the method, high-precision and low-calculation-overhead target detection can be realized in a complex scene, and meanwhile, the detection performance and robustness are remarkably improved.

The invention relates to a model optimization method and device for a low-resource scene, electronic equipment and a medium, and the method comprises the steps: obtaining a pre-training model, the pre-training model being a Transform model, for any linear layer in the Transform model, connecting a pair of low-rank matrixes in parallel in the linear layer to form a LoRA increment path, and freezing a backbone network in the pre-training model; for any linear layer in the Transform model, a learnable gate is added on a LoRA increment path, and an improved LoRA module is formed; and obtaining training data, training the pre-training model according to the training data to obtain a target model, and performing semantic analysis on the to-be-processed text according to the target model. According to the method provided by the invention, through collaborative design of structural innovation and an optimization mechanism, the performance and stability of the efficient parameter fine tuning method in a low-resource scene are remarkably improved, and the method has a good industrial application prospect.

The invention relates to a distributed optical fiber sound wave sensing data feature extraction method and device, and the method comprises the steps: carrying out the feature extraction of distributed optical fiber sound wave sensing data through a Transform network model which is pre-trained through a mask reconstruction task; the Transform network model is improved by utilizing an attention mechanism and comprises a coding network and a decoding network which are the same in structure, each of the coding network and the decoding network comprises a linear layer, a position coding layer and a multi-layer cascaded converter encoder, and the system is used for realizing the method. Compared with the prior art, the method does not need to use data labels for model training, has excellent generalization ability, and can be suitable for various tasks based on distributed optical fiber sound wave sensing data.

The invention relates to a traffic flow prediction method based on temporal-spatial feature fusion Transform in an edge environment, and the method comprises the steps: carrying out the feature embedding and coding of original traffic flow data; then, constructing a multi-head convolution low-rank decomposition attention mechanism to capture a long-term time dependency relationship and obtain local context information; then, constructing an attention map convolution to capture a spatial dependency relationship; and finally, performing adaptive fusion on the spatial-temporal characteristics through a gating unit, and further realizing accurate prediction of future traffic flow by using a feedforward neural network and a linear layer. The method can improve the prediction precision, effectively reduce the resource overhead, and improve the prediction efficiency.

The invention relates to the technical field of satellite altimetry, in particular to an altimetry method based on a residual network multi-mode deep learning model, which can remarkably improve the robustness of SSH inversion under sea condition noise interference, and is characterized in that a fusion self-attention mechanism and a residual network multi-mode deep learning model ViTResNetMDL is established, and the model ViTResNetMDL is applied to improve the robustness of the SSH inversion under sea condition noise interference. A satellite-borne GNSS-R sea surface height inversion geometric physical model is established, a ViTResNetMDL model is composed of ResNet50, ViT and MHL-NN, the first part adopts a ResNet50 convolution structure to extract effective features from an original DDM, the second part adopts a ViT layer which captures global features in an effective scattering area DDM based on a self-attention mechanism in a Transform module, the third part is composed of linear layers, and the fourth part is composed of linear layers. And using the MHL-NN to retrieve the sea surface height.

The invention discloses an aircraft target trajectory prediction method and device based on multi-scale Mama, and belongs to the field of flight trajectory prediction, and the method comprises the steps: S1, carrying out the data preprocessing of trajectory sequence data, and recording the processed trajectory sequence as Xt: t + n; s2, converting Xt: t + n into a track sequence T which learns a change relation of time steps; s3, embedding the T code into a high-dimensional space, and capturing an interdependency relationship S of different components in a high-dimensional embedded vector; s4, compressing S at different scales, extracting features of different resolutions, and extracting to obtain a combined feature M of the two features; and S5, learning attention distribution of the joint feature M by using a wavelet attention module, and restoring a historical trajectory sequence Yt: t + n and a predicted trajectory point Yt + n + 1 through a linear layer and inverse wavelet transform. According to the method, high prediction precision and high response speed are balanced at the same time, and reliable technical support is provided for airspace management.

The embodiment of the invention discloses a quantitative model degradation correction method and device, electronic equipment and a storage medium, relates to the technical field of data processing, and is convenient for reducing errors generated by a quantitative model. The method comprises the following steps: acquiring a key operator of an original model; implementing the key operator by adopting a model reasoning framework to obtain a model reasoning operator of a quantized version; registering the model reasoning operator into a deep learning training framework, and replacing a corresponding operator in the deep learning training framework with the model reasoning operator to obtain a quantitative model; injecting a fine-tuning low-rank weight into each linear layer of the quantitative model to obtain an injected quantitative model; performing fine tuning on the post-injection quantification model to obtain a fine tuning parameter result; and loading the fine tuning parameter result to the quantitative model. The method is used for degeneration correction of the quantitative model.

The invention discloses a model training method and device, equipment, a storage medium and a program product, and the method comprises the steps: carrying out the type conversion of an input tensor of a second data type based on a zooming parameter corresponding to a linear layer in a forward process, and obtaining an input tensor of a first data type; after the calculation of the linear layer is completed, converting the output tensor of the first data type into an output tensor of a second data type; in a reverse process, performing type conversion on the model parameter of the first data type based on a scaling parameter corresponding to a linear layer to obtain a model parameter of a second data type; determining a gradient of the second data type based on the model parameter of the second data type and the loss value of the second data type; and updating model parameters of the model by using the gradient of the second data type to obtain a trained model. Therefore, under the condition that the model precision is guaranteed, the model training efficiency is improved, and the requirements for computing power and storage during model training are reduced.

The invention discloses a dynamic scene re-operation mirror video generation method and system based on a diffusion model, and belongs to the technical field of computer vision and video generation. A diffusion generation architecture with a control branch is adopted, and the core is composed of an embedded layer, a main branch and a control branch. In the control branch, the output of each sub-block is added with the output of the corresponding block of the main branch after being processed by the zero initial linear layer, and the sum is input into the next block of the main branch. During training, generating a rendered video by using the target video and the reference video in the same scene; and inputting the target video latent variable after noise addition into the control branch, inputting the splicing result of the target video, the reference video and the rendering video latent variable into the main branch, and simultaneously providing the text latent variable of the reference video for the two branches as a condition. During generation, the model finally generates a target video latent variable through step-by-step denoising and decodes the target video latent variable into a target track video, and it is ensured that the motion of a moving object in a scene of a generated video and a reference video is consistent at the same time.

The invention discloses a data processing method and system for a multi-head potential attention model based on TTD compression and a terminal, and the method comprises the steps: constructing a large language model, and processing a plurality of linear layers in the large language model in a TTD compression and decomposition mode, thereby reducing the weight number in the model, and improving the data processing efficiency. And on a hardware level, targeted optimization is carried out on a data stream subjected to linear layer reasoning after TTD compression, so that a final model reasoning result is output. According to the method, a multi-head potential attention structure in a large language model is converted, so that the requirement for KV cache during model reasoning is reduced, the weight number is reduced, the long text output capacity of the model on edge equipment is improved, matrix calculation in the reasoning process is carried out subsequently by adopting a calculation structure of a group vector systolic array, and the reasoning efficiency is improved. And limited hardware resources are efficiently utilized.

The invention discloses a time series prediction method based on a periodic quantum machine learning network, and the method comprises the steps: carrying out the normalization and dimension reduction processing of time series data through a data preprocessing module; secondly, the classical data is converted into a quantum state through quantum state coding; then, periodic features in the time sequence are extracted through a periodic quantum block module, and the module is composed of a rotating door and a control door and can effectively simulate Fourier series; then, the quantum state is converted back to classical data through quantum measurement; and finally, outputting a final prediction result through a data post-processing module in combination with residual connection and a linear layer. According to the method, the advantages of quantum calculation are utilized, the periodic characteristics in the time sequence are captured through the data re-uploading quantum circuit, and the prediction accuracy and the calculation efficiency are remarkably improved. The method is superior to a traditional method in data synthesis and real world time sequence prediction tasks, and shows high generalization ability in non-periodic tasks.

The present invention relates to methods and systems for improving the efficiency of artificial neural networks by configuring them to implement linear dynamical systems that compute state updates in linear time. More specifically, the present invention specifies methods and systems for setting the weights of least one linear artificial neural network layer by (a) selecting an set of basis vectors that define a desired impulse response for the layer, (b) deriving a matrix that produces this desired impulse response over a target temporal window, (c) deriving a matrix that dampens the impulse response to zero outside of the target window, and (d) combining these two matrices together through addition to obtain the layer's recurrent weights. Systems composed of at least one such linear layer are applied to a time series of input data elements to produce outputs that encode the results of pattern classification, signal processing, data representation, and data generation tasks.

The invention discloses a ground penetrating radar C-scan inversion method and system. The method comprises the steps that XZ slice data, YZ slice data and XY slice data of a to-be-inverted target are acquired and preprocessed to construct a training data set; constructing a ground penetrating radar slice data inversion initial model based on a convolutional layer, a linear layer, a Patch Merging layer, a Patch Expanding layer, a projection layer, residual connection and jump connection, and training to obtain a ground penetrating radar slice data inversion model; inputting detection data obtained in the actual three-dimensional detection scene into the ground penetrating radar slice data inversion model for processing to obtain inversion slice data; and performing splicing and three-dimensional reconstruction on the obtained inversion slice data to complete C-scan inversion of the ground penetrating radar. According to the method, the detection data of the XZ slice, the YZ slice and the XY slice are processed by adopting the constructed inversion model, and splicing and three-dimensional reconstruction are combined, so that the C-scan inversion of the ground penetrating radar is realized, the reliability is higher, the accuracy is better, and the efficiency is higher.

The invention discloses a video understanding large model optimization and evaluation method, system and device and a storage medium, which are in one-to-one correspondence, and in the scheme, a new connector structure is designed to improve the global time sequence understanding capability, and the new connector structure comprises a space-time downsampler, a local bidirectional Mama structure and a linear layer. The time-space downsampler can reduce the token storage overhead; meanwhile, the local bidirectional Mama structure makes up for the problem that the receptive field is limited on one hand, and on the other hand can model intra-frame features and inter-frame features at the same time; in addition, training of the connector is low in cost, a three-stage progressive training strategy is used, and catastrophic forgetting can be resisted. Moreover, a semi-automatic data generation pipeline is constructed, and global time sequence understanding data is provided based on the pipeline, so that the evaluation vacancy of the existing reference field in the capability is made up.

The invention belongs to a point cloud reconstruction method, and provides a power equipment three-dimensional point cloud reconstruction method and a related device for solving the technical problems that the accuracy of reconstructed and restored object form is low and key features are difficult to capture during detail restoration in two common curved surface reconstruction methods at present. And inputting into a three-dimensional point cloud reconstruction model of the power equipment to obtain a reconstruction result. Wherein the calculation method in the three-dimensional point cloud reconstruction model of the power equipment comprises local feature weighting, global feature fusion and occupancy prediction, the local feature weighting is based on the similarity of query point and neighborhood point features, the global feature fusion is based on the global feature weighting, and the occupancy prediction is realized through linear layer decoding. The detail reconstruction precision is effectively improved, and the accuracy of overall reconstruction is guaranteed.

The invention relates to an NDVI product generation method and device based on a diffusion model, equipment and a medium, and relates to the technical field of remote sensing data processing, and the method comprises the steps: obtaining a plurality of NDVI images corresponding to a target time period; each NDVI image is input into a pre-constructed diffusion generation model, and an NDVI product corresponding to the NDVI image is generated; wherein the diffusion generation model which is constructed in advance comprises a Stable Diffusion model and a LoRA fine tuning module; the Stable Diffusion model comprises a VAE module and a diffusion module constructed on the basis of a U-Net framework; and the LoRA fine tuning module is embedded in a Query linear layer, a Key linear layer and a Value linear layer in the attention structure. The method has the technical effect of effectively generating the high-time-resolution NDVI product by using the diffusion model.

The invention discloses an internal threat detection method based on a Transform-DCN model, relates to the field of artificial intelligence and network security, and solves the problems of insufficient time sequence feature relevance and insufficient abnormal mode capture in existing detection. Comprising the following steps: extracting single-day features of user behaviors through multi-dimensional data preprocessing, and constructing a multivariable time sequence; the method comprises the following steps: inputting time sequence data into a Transform-DCN model for detection, firstly, improving feature dimensions by an embedded layer through linear transformation, fusing position codes, and capturing absolute time sequence information; then analyzing a global dependency relationship through a multi-head self-attention mechanism of a Transform module, and detecting context exception; meanwhile, extracting local abnormal features by using deformable convolution (DCN); and finally, fusing the global features and the local features by an output layer, and outputting whether a threat behavior is a threat behavior through a linear layer. According to the method, the accuracy and integrity of internal threat detection can be effectively improved, and the purpose of reducing network security risks of organizations and enterprises is achieved.

The present invention discloses an image layout attribute prediction method based on a graph neural network, comprising the following steps: a) determining various attributes to be predicted for the image layout; preprocessing the image and converting it into a graph structure including nodes and edges; b) constructing network blocks; which consist of multiple layers. In each layer, first, it passes through a linear layer, then through a multi-head graph convolutional network, and finally through a two-layer multi-layer perceptron to further enhance the diversity of feature representations; c) for the image processed by multiple layers of such network blocks, the number of output features decreases layer by layer, and the network has a pyramid structure. After several layers, the final feature information of each node can be obtained. d) Finally, these features are input into a fully connected network and a classifier to obtain the predicted values of various layout attributes of the image. Compared with the existing methods, the present invention has strong feature extraction and fusion capabilities, and can improve the accuracy of predicting the image layout to a certain extent.

The invention discloses a classification fault detection method based on a graph neural network, and the method comprises the steps: inputting time series data, projecting the input data to a set dimension through a linear layer, employing a two-layer one-dimensional convolutional neural network to extract time features, dynamically generating an adjacent matrix through a radius according to the Euclidean distance, and based on the adjacent matrix, in each layer of graph convolution, carrying out the classification fault detection of each layer of graph convolution. The method comprises the following steps of: performing spatial aggregation on adjacent node features, learning the weight of important nodes by using an attention mechanism, repeating three-layer GATv2 convolution, adding residual connection and normalization processing into output of each layer, finally obtaining node feature representation of spatial aggregation, and finally converting extracted spatial and temporal features into classification results through a decoder to detect faults; the method solves the problem that the prior art cannot be competent for multi-source data and diversified features in a modern industrial system, combines a memory one-dimensional convolutional neural network and a GNN deep learning method, can extract spatial and temporal features at the same time, and is significantly superior to a traditional statistical method in fault detection accuracy.

The invention discloses a lane line detection method based on SDAMamba and FFCDConv, and relates to the field of lane line detection. The method comprises the following steps: preprocessing a training set lane line image to obtain an updated training set; improving the DAMamba model to obtain an SDAMamba model for extracting multi-scale features of the updated training set image; extracting depth multi-scale features by using a feature pyramid neck network; mapping the depth multi-scale feature to a Hough parameter space by using DHT and selecting a Hough feature corresponding to a lane; jointly inputting the Huffcharacteristics and the depth multi-scale characteristics into an FFCDConv module, predicting the characteristics of different lane instances, and obtaining a predicted lane position map through a lane decoder; and inputting the predicted lane position map into a linear layer to determine a final lane position, and obtaining a lane line detection image. According to the method, the problem of breakage easily occurring in a lane line detection result can be repaired in a targeted manner, and the confusion rate of adjacent lanes is remarkably reduced.

The invention relates to a model training method and device, a storage medium and a program product. The method comprises the following steps: in a forward propagation stage, for any check point module, storing input and output of the check point module in a video memory, and releasing an intermediate activation value of the check point module in the video memory; the input of the check point module is used for recalculating an intermediate activation value, and the output of the check point module is used for forward calculation of subsequent modules of the check point module; and in a back propagation stage, for any check point module, responding to the condition that the last layer of the check point module is a linear layer, skipping forward calculation of the last layer, calculating a gradient according to a gradient formula corresponding to the last layer, and completing back propagation of the check point module according to the gradient of each layer in the check point module. According to the method and the device, the calculation cost can be remarkably reduced while the calculation precision and the video memory saving amount are the same as those of a standard re-calculation scheme.

The invention discloses a robot action sequence generation method based on a hierarchical attention mechanism, and the method comprises the steps: 1, inputting a plurality of robot operation images into a CNN layer, and extracting an image feature sequence; projecting the multiple pieces of joint position information into a joint state vector sequence through a linear layer of the CNN layer; taking a to-be-executed action sequence as a prediction target; taking the joint state vector sequence, the image feature sequence and the prediction target as input data of an encoder; 2, mapping the joint state vector sequence into a style variable by adopting an encoder, and adding randomly initialized learnable vector marks at the front ends of the image feature sequence, the joint state vector sequence and the action sequence to be executed; and 3, based on the sequence sparse attention, the trained decoder is adopted to predict a future action sequence of the mechanical arm of the robot. The invention provides a new solution for improving the fine operation capability of the robot.

The invention relates to a password side channel analysis method based on a secondary start genetic algorithm, and belongs to the technical field of information security passwords. The method comprises the following steps: firstly, constructing a second-order leakage corresponding formula of a cryptographic algorithm intermediate value and a leakage waveform for a first-order mask protection scheme realized by nonlinear layer table look-up; then, aiming at the generated second-order leakage, extracting an intermediate value of bit multiplexing in a register to construct an offset calculation formula of the intermediate value and a typical value; and finally, constructing a genetic algorithm fitness calculation formula for algorithm intermediate value offset, performing key recovery by adopting a secondary start genetic algorithm heuristic learning method for reducing the number of individuals in a dynamic population, and performing side channel analysis on a lightweight cryptographic algorithm for realizing a nonlinear layer by parallel table look-up.

The invention provides a Bayesian nerve radiation field modeling method and system based on uncertainty perception and dynamic importance sampling. The method comprises the following steps: replacing a full connection layer in a multi-layer perceptron with a Bayesian linear layer to obtain a Bayesian neural radiation field BN-NeRF model; acquiring a data set containing house source photos of different viewing angles and corresponding camera positions, and training a BN-NeRF model by using the data set; performing preliminary coarse sampling on each light passing through the house source scene to obtain a coarse sampling point set, and performing uncertainty evaluation on each sampling point in the coarse sampling point set by adopting a trained BN-NeRF model; according to the uncertainty evaluation result corresponding to the preliminary coarse sampling, performing secondary sampling on each light passing through the housing resource scene to obtain a fine sampling point set; and integrating the coarse sampling point set and the fine sampling point set to generate a final sampling point set, calculating the color and volume density of each sampling point in the final sampling point set by adopting a trained BN-NeRF model so as to carry out volume rendering, and generating a final house viewing picture.

A multi - meteorological element data prediction method based on fine - tuning training of a deep neural network, including pre - processing a multi - meteorological data set to obtain a preliminary prediction result, and splitting the preliminary prediction result by meteorological elements or by time steps; respectively matching each time series with a general lightweight fine - tuning module for fine - tuning, and fully extracting features through a multi - layer perceptron composed of multiple linear layers; multiplying the output of the linear layer by the weight V obtained by layer normalization to obtain the prediction result corresponding to this lightweight fine - tuning module; finally, combining the results of each lightweight fine - tuning module. By introducing advanced time - series analysis techniques and a dynamic adjustment mechanism, the present invention can monitor the model performance in real time and automatically adjust when necessary, thereby reducing the accumulation of prediction errors, maintaining the long - term stability and reliability of the model, being applicable to the prediction of any meteorological element expressed in time series, and allowing users to independently select a suitable pre - trained deep - learning model.