4 results about "Global statistics" patented technology

Mechanisms are provided for dynamically modifying maximum heap sizes of local memory heaps. A plurality of applications execute that each have a corresponding local garbage collector in a plurality of local garbage collectors. A global garbage collection manager executes and communicates with the local garbage collectors. The global garbage collection manager maintains global statistics information that includes statistics information communicated to it from each of the local garbage collectors. A first local garbage collector, associated with a first application, communicates first statistics information of the first application, in response to a first local garbage collection operation executed by the first local garbage collector, to the global garbage collection manager. The global garbage collection manager dynamically modifies a heap size associated with the first application based on the first statistics information and the global statistics information.

In a database system, wherein data is stored as objects within an object storage system, a system and method for estimating object cardinality, determining query execution plan costs, and selecting a query plan for execution by the database system. Multiple object cardinality estimation approaches for estimating the number of objects to be accessed for a given query condition on a column of a relation composed of a set of objects, where each object maintains the minimum value and the maximum value of individual columns are presented. A set of global statistics is also maintained, consisting of the total number of objects and the minimum and maximum values of individual columns. The object cardinality estimation is determined based on the global statistics without retrieving individual object-level statistics.

This invention discloses an euploidy prediction method based on the fusion of blastocyst developmental dynamics and spherical harmonic decomposition. The method acquires multifocal plane image sequences of blastocysts using a time-difference imaging system and extracts dynamic parameters. After semantic segmentation and depth estimation, a three-dimensional surface model is obtained through surface reconstruction. The trophoblast cell instances are segmented to obtain three-dimensional centroid coordinates. These centroid coordinates are radially projected onto a unit sphere, and a spherical density function is constructed after excluding the inner cell mass mask. Spherical harmonic decomposition is then normalized using Monte Carlo zero-model normalization to obtain normalized power spectra at each degree. The basic morphological features, dynamic parameters, and spherical harmonic features are fused and filtered before training an ensemble learning model to output prediction results. This invention is the first to introduce spherical harmonic analysis into blastocyst assessment, capturing multi-scale spatial distribution information that global statistics cannot obtain. The power spectrum exhibits rotational invariance, and zero-model normalization eliminates cell number confounding.

The application discloses a two-stage speculative decoding optimization method and system for large model inference acceleration. First, a general domain dataset and a specific task domain distillation dataset are constructed, and a frozen verification model and a candidate generation model to be trained are established. Based on the general domain dataset, the first-stage preliminary capability training of the candidate generation model is performed, so that the candidate generation model enters the initial interval effectively aligned with the verification model. Based on the task domain distillation data, the second-stage key mark optimization training is performed. The ratio of the verification model probability to the candidate generation model probability is calculated for each to-be-predicted mark to obtain the potential contribution index of the candidate generation model. The dynamic screening threshold and the sparse mask are generated by using global statistics, and the loss optimization is only performed on the key contribution mark. Finally, the trained candidate generation model and the verification model are used to perform speculative decoding inference. The method improves the acceptance length of the candidate mark and the inference acceleration effect, and reduces the training cost.