Encoding and Adaptive, Scalable Accessing of Distributed Models
First Claim
Patent Images
1. A system comprising:
- a plurality of machine translation resource servers, each machine translation resource server storing and operable to serve a partition of a collection of machine translation resource data for translation from a source language to a target language, the respective partitions together constituting the collection of machine translation resource data and each respective partition being less than the collection of machine translation resource data; and
at least one translation server operable to receive source text in the source language to be translated into the target language, the translation server further operable to obtain machine translation resource data from the plurality of machine translation resource servers and to use the obtained machine translation resource data to translate the source text into the target language.
2 Assignments
0 Petitions
Accused Products
Abstract
Systems, methods, and apparatus for accessing distributed models in automated machine processing, including using large language models in machine translation, speech recognition and other applications.
-
Citations
91 Claims
-
1. A system comprising:
-
a plurality of machine translation resource servers, each machine translation resource server storing and operable to serve a partition of a collection of machine translation resource data for translation from a source language to a target language, the respective partitions together constituting the collection of machine translation resource data and each respective partition being less than the collection of machine translation resource data; and at least one translation server operable to receive source text in the source language to be translated into the target language, the translation server further operable to obtain machine translation resource data from the plurality of machine translation resource servers and to use the obtained machine translation resource data to translate the source text into the target language. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53)
-
-
54. A system comprising:
-
a translation server operable to perform machine translation obtaining translation model data from a translation model for translation between a source language and a target language and language model data from a language model for the target language, the translation server further operable to translate text in the source language into the target language using the obtained translation model data and language model data, the translation server comprising; a request queue operable to store requests for language model data to be obtained for translating a segment in the source language, and a segment translation server cache operable to store language model data obtained by the requests by the translation server. - View Dependent Claims (55, 56, 57, 58, 59, 60, 61, 62, 63, 64)
-
-
65. A method, comprising:
-
dividing a collection of machine language translation resource data for translation from a source language to a target language into a plurality of partitions each being less than the collection of machine language translation resource data; storing the plurality of partitions on different computer servers, respectively; and operating a machine translation server to access and use the collection of machine language translation resource data on the different computer servers to perform translation from the source language into the target language. - View Dependent Claims (66, 67, 68, 69, 70, 71, 72, 73, 74)
-
-
75. A method for machine translation of text from a source language into a target language using a translation model for translation between the source language and the target language and a language model for the target language, comprising:
-
partitioning the translation model into a plurality of partitions of different data, wherein each translation model partition is less than the translation model; storing the plurality of translation model partitions on a plurality of different translation model servers; partitioning the language model into a plurality of language model partitions of different data, wherein each language model partition is less than the language model; storing the plurality of language model partitions on a plurality of different language model servers; monitoring work load of a plurality of translation severs each operable to translate text in the source language into the target language using the translation model and the language model; distributing segments of a text to be translated from the source language into the target language to one or more selected translation servers from the plurality of translation servers based on the work load; operating each selected translation server to access the translation model severs and the language model servers to fetch desired translation model data and language model data for each respective segment to be translated; and compiling translated segments from the selected translation servers to produce a translated text. - View Dependent Claims (76, 77, 78, 79)
-
-
80. A computer implemented method, comprising:
-
receiving a client document in a source language to be translated into a target language; dividing the client document into segments to translate each segment; accessing at least one of a plurality of different language model servers, which collectively store a language model for the target language, to retrieve selected language model data related to translation one of the segments, each language model server storing and operable to serve a partition of the language model; and translating the segment into the target language using the retrieved selected language model data. - View Dependent Claims (81, 82, 83, 84, 85, 86, 87, 88)
-
-
89. A method comprising:
-
is using a machine translation system to receive text in a source language from a client and to translate the text into a target language, wherein the translating in the machine translation system comprises; selecting a portion of the text to translate at a low translation quality to produce an initial translated portion while translating the selected portion at a high translation quality; delivering the initial translated portion to the client while continuing translating the selected portion at the high translation quality; and after the selected portion is translated into a second translated portion at the high translation quality, delivering the second translated portion at the high translation quality to the client to automatically replace the initial translated portion. - View Dependent Claims (90, 91)
-
Specification