Search Results for "method"

...These models expand and optimize the Chinese vocabulary on the basis of the original Llama-2, use large-scale Chinese data for incremental pre-training, and further improve the basic semantics and command understanding of Chinese. Performance improvements. The related model supports FlashAttention-2 training, supports 4K context and can be extended up to 18K+ through the NTK method.

OpenNN is a software library written in C++ for advanced analytics. It implements neural networks, the most successful machine learning method. Some typical applications of OpenNN are business intelligence (customer segmentation, churn prevention…), health care (early diagnosis, microarray analysis…) and engineering (performance optimization, predictive maitenance…). OpenNN does not deal with computer vision or natural language processing. The main advantage of OpenNN is its high performance. ...

...Implements a number of meta-heuristic algorithms for Non-Linear Programming, including Genetic Algorithms, Differential Evolution, Evolutionary Algorithms, Simulated Annealing, Particle Swarm Optimization, Firefly Algorithm, Monte-Carlo Search, Local Search algorithms, Gradient-Descent-based algorithms, as well as some well-known network flow and other graph algorithms. A fast parallel implementation of the network simplex method, and some full-fledged parallel/distributed MIP solvers will be added in the next version. In general, emphasis is given in improving the efficiency of the algorithms in shared-memory models via java threads, since multi-core machines are so wide-spread today.

Maximum entropy is a powerful method for constructing statistical models of classification tasks, such as part of speech tagging in Natural Language Processing. Several example applications using maxent can be found in the OpenNLP Tools Library.