Sequential sampling from multisets
AssigneeMicrosoft Technology Licensing, LLC
InventorJeffrey William Pasternack
Life of this patent
Estimating the term…
Abstract
Systems, devices, media, and methods are presented for dynamically sampling training elements while training a plurality of machine learning models. The systems and methods access a plurality of training elements and generate an element buffer including a specified number of positions. The systems and methods generate a random value for each position for the element buffer and map a training element of the plurality of training elements to each position within the element buffer in a first order. The systems and methods generate a second order for the training elements and allocate the training elements to a plurality of machine learning models according to the second order.
Claims20
A method comprising:
- accessing a plurality of training elements;
- generating an element buffer including a specified number of positions;
- generating a random value for each position of the element buffer;
- based on the random value for each position, mapping a training element of the plurality of training elements to each position within the element buffer, the training elements mapped to positions of the element buffer in a first order;
- generating a second order for the training elements by modifying positions of the training elements in the element buffer;
- and
- allocating the training elements in the element buffer to a plurality of machine learning models according to the second order for generation of models related to natural language processing.
A system comprising:
- one or more processors;
- and
- a processor-readable storage device comprising processor executable instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:
- accessing a plurality of training elements;
- generating an element buffer including a specified number of positions;
- generating a random value for each position for the element buffer;
- based on the random value for each position, mapping a training element of the plurality of training elements to each position within the element buffer, the training elements mapped to positions of the element buffer in a first order;
- generating a second order for the training elements by modifying positions of the training elements in the element buffer;
- and
- allocating the training elements in the element buffer to a plurality of machine learning models according to the second order for generation of models related to natural language processing.
A processor-readable storage device comprising processor executable instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:
- accessing a plurality of training elements;
- generating an element buffer including a specified number of positions;
- generating a random value for each position of the element buffer;
- based on the random value for each position, mapping a training element of the plurality of training elements to each position within the element buffer, the training elements mapped to positions of the element buffer in a first order;
- generating a second order for the training elements by modifying positions of the training elements in the element buffer;
- and
- allocating the training elements in the element buffer to a plurality of machine learning models according to the second order for generation of models related to natural language processing.
Dependencies are inferred from the claim text.