US Patent Application 18219555. DETERMINISTIC TRAINING OF MACHINE LEARNING MODELS simplified abstract
Contents
DETERMINISTIC TRAINING OF MACHINE LEARNING MODELS
Organization Name
Inventor(s)
Gaurav Mishra of Sunnyvale CA (US)
Adam Joseph Roberts of Durham NC (US)
Noam M. Shazeer, Jr. of Palo Alto CA (US)
Maarten Paul Bosma of Cupertino CA (US)
DETERMINISTIC TRAINING OF MACHINE LEARNING MODELS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18219555 titled 'DETERMINISTIC TRAINING OF MACHINE LEARNING MODELS
Simplified Explanation
The patent application describes a method for training a machine learning model using a deterministic data pipeline. Here are the key points:
- The method involves generating a deterministic training dataset by transforming raw training examples obtained from a raw data source into pre-processed training examples.
- Each pre-processed training example is assigned a unique index.
- The pre-processed training examples are cached into a specified cache directory.
- The method also involves using the deterministic training dataset to train a machine learning model.
- A second request is received, specifying a start index for reading the pre-processed training examples from the cache directory.
- The pre-processed training examples with indices beginning from the start index are read from the cache directory.
- The read training examples are provided in the order of the assigned indices for use in training the machine learning model.
Original Abstract Submitted
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a machine learning model using a deterministic data pipeline. One of the methods may include receiving a first request to generate a deterministic training dataset: transforming raw training examples obtained from the raw data source into pre-processed training examples; assigning a unique index to each pre-processed training example; and caching the pre-processed training examples into the cache directory specified in the received first request; receiving a second request to use the deterministic training dataset to train a machine learning model, the second request specifying a start index; and in response to receiving the second request: reading, from the cache directory, the pre-processed training examples that have indices beginning from the start index; and providing the read training examples in an order of the assigned indices for use in training the machine learning model.
