Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This action is in response to the original filing on 11/19/2023. Claims 1 – 20 are pending and have been considered below.
Claim Objections
Claims 2 – 9, 11 – 18 and 20 are objected to because of the following informalities:
“ of Claim” Lowercase “c” in claims 2 – 9, 11 – 18 and 20.
Appropriate corrections are required.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitations uses the term “means” coupled with functional language without reciting sufficient structure, material, or acts for performing the claimed function. Such claim limitations are: means for: receiving data from an application executing on a host device to store the received data; in claim 19, and means for: receiving a query for information associated with particular data stored in the secondary storage; in claim 20.
Because these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitations recites sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1 - 4, 6, 7, 10 - 13, and 15 - 20, are rejected under 35 U.S.C. 103 as being unpatentable over Bedadala et al. (US 20190250839 A1) (hereinafter Bedadala), in view of Mishra et al. (US 20250045534 A1) (hereinafter Mishra).
Regarding claim 1, Bedadala teaches A method performed by a storage system including a primary storage and a secondary storage, the method comprising (Bedadala, [FIG. 1D],) Examiners Remarks: displays a system including the, a primary and secondary storage.
PNG
media_image1.png
601
584
media_image1.png
Greyscale
receiving data from an application executing on a host device of the storage system to store the received data (Bedadala,[0006], “retrieve data object usage data [receiving data from an application] associated with the client computing device [a host device]”, furthermore “the client computing device to store one or more data objects in a secondary storage device”. Examiner Remarks: wherein the data object is data associated with an application, that is being executed on a client computing device (the host device) to be stored.);
determining whether the received data is to be stored in the secondary storage or in the primary storage based at least in part on the determination on whether the received data is to be used {for finetuning the LLM1,} (Bedadala,[0040], ” a storage policy generally comprises a data structure or other information source that defines a set of preferences or other criteria for performing information management operations. Storage policies can include one or more of the following: (1) what data will be associated with the storage policy; (2) a destination to which the data will be stored;”, furthermore “determine which data to transfer from a primary storage device to a secondary storage device.” Examiners Remarks: Storage policy determines; the criteria’s of the data, i.e. if the received data is associated with a policy, the destination of the data, therefore under BRI data can be stored in primary storage or secondary storage based at least on a determination of whether it is used.) wherein stored data is accessed faster by the host device from the primary storage than from the secondary storage (Bedadala,[0093], “Primary storage devices 104 storing primary data 112 may be relatively fast and/or expensive technology (e.g., flash storage, a disk drive, a hard-disk storage array, solid state memory, etc.), typically to support high-performance live production environments.” Furthermore Bedadala,[0191], “movement of data from primary storage devices 104 to secondary storage devices 108, or between tiers of secondary storage devices 108. With each tier, the storage devices may be progressively cheaper, have relatively slower access/restore times, etc. For example, movement of data between tiers may occur as data becomes less important over time.” Examiners remarks: A primary storage, ( first tier of storage device(s)) are faster and more accessed than a secondary storage ( second tier of storage device(s)));
{and in response to determining that the received data is to be used for fine-tuning the LLM2} storing the received data in the secondary storage (Bedadala,[0319], “data objects to store in a secondary storage device”).
Bedadala does not expressly teach determining whether the received data is to be used for fine-tuning a Large Language Model (LLM); for fine-tuning the LLM, and in response to determining that the received data is to be used for fine-tuning the LLM,
However, Mishra teaches determining whether the received data is to be used for fine-tuning a Large Language Model (LLM) (Mishra, [0004], “ it can be assumed that each subsequent response is an improvement on the preceding response. Some of these implementations described herein relate to using the described techniques at inference time, to provide an improved response to a user that is responsive to the user's NL based input”, Furthermore, “the described techniques to generate training data by storing the NL based input along with an improved response that is responsive to the NL based input as an instance of training data. This training data can be used to fine-tune an LLM,” wherein in the determination for fine-tuning the LLM is the subsequent NL based input, and if there is no subsequent NL input there is not fine-tuning of the LLM ); for fine-tuning the LLM (Mishra, [0004], “it can be assumed that each subsequent response is an improvement on the preceding response.” Furthermore Mishra, [0016], ” generate synthetic training data for fine-tuning an LLM for subsequent utilization” Examiner Remarks: wherein in for fine-tuning the LLM, the subsequent improved responses from the NL based input data are the fine-tuning of the LLM).
and in response to determining that the received data is to be used for fine-tuning the LLM, (Mishra, [0004], “ it can be assumed that each subsequent response is an improvement on the preceding response. Some of these implementations described herein relate to using the described techniques at inference time, to provide an improved response to a user that is responsive to the user's NL based input”, Furthermore, “the described techniques to generate training data by storing the NL based input along with an improved response that is responsive to the NL based input as an instance of training data. This training data can be used to fine-tune an LLM,” Examiner Remarks: wherein in the determination for fine-tuning the LLM is the subsequent NL based input, and if there is no subsequent NL input there is not fine-tuning of the LLM ).
Because Bedadala and Mishra are analogous art and within the same field of endeavor, specifically computer-implemented systems for the efficient utilization of large language models in storage systems. Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine Bedadala’s, machine learning based on data storage with Mishra’s, efficient fine-tuning of large language models (LLM). This modification would have been motivated by the desire to train a large language model (LLM) for performing specific task and reducing usage of computational resources. (Mishra, [0002].)
Regarding claim 2, the combination Bedadala and Mishra teaches the claimed invention as claimed in The method of Claim 1, further comprising:
receiving a query for information associated with particular data stored in the secondary storage (Bedadala,[0007], “obtain a request [receiving a query] to store in the secondary storage device a first data object in the one or more data objects from the client computing device, where the client computing device generates the request in response [associated with particular data] to a prediction produced by the data storage ML model”); and
inputting the query into the LLM to provide the information without accessing the particular data. (Mishra, [0025], “generating a response [to provide the information] to a particular NL based input [inputting the query into the LLM], the LLM leverages the underlying data [the particular data.] on which it was trained. In this way, an LLM can be considered to be a database structure with information stored in the parameters of the LLM.” Examiners Remarks: Wherein the LLM provides the information without access the particular data or dataset of the LLM ).
Regarding claim 3, the combination Bedadala and Mishra teaches the claimed invention as claimed in The method of Claim 1, further comprising:
receiving a query for information (Mishra, [0041], “ NL based input described herein can be a query”, furthermore, Mishra, [0007], “, the NL based input, information associated with a user, and/or other context data such as the application used to enter the NL based input”);
determining that the query for information is not associated with a data type used to fine-tune the LLM (Mishra, [0005], “it can be determined that further improvement of a response is not necessary. For instance, if the appropriate response to an NL based input is either “yes” or “no”, or if it is determined that a particular response (e.g., prior to the final response) is correct, then no improvement in the subsequent responses is required. In some of these cases, each subsequent response can simply be a repetition of the “correct” response. In other of these cases, the LLM can halt further processing of the NL based input”)
Regarding claim 4, the combination Bedadala and Mishra teaches the claimed invention as claimed in The method of Claim 1, further comprising determining whether the received data is to be used for fine-tuning the LLM (Mishra, [0004], “ it can be assumed that each subsequent response is an improvement on the preceding response. Some of these implementations described herein relate to using the described techniques at inference time, to provide an improved response to a user that is responsive to the user's NL based input”, Furthermore, “the described techniques to generate training data by storing the NL based input along with an improved response that is responsive to the NL based input as an instance of training data. This training data can be used to fine-tune an LLM,” Examiner Remarks: wherein in the determination for fine-tuning the LLM is the subsequent NL based input, and if there is no subsequent NL input there is not fine-tuning of the LLM )
based on at least one characteristic of the received data. (Bedadala,[0004], “A media agent using the monitored data usage data to train a data storage machine learning model” furthermore, “The data storage machine learning model may be trained such that the model, given an input such as time, an identification of data object(s), data object metadata, etc ” )
Regarding claim 6, the combination Bedadala and Mishra teaches the claimed invention as claimed in The method of Claim 1, wherein the primary storage includes one or more first storage devices of a first type (Bedadala, [0093], “Primary storage devices 104 storing primary data” Examiner Remarks: Wherein Primary storage devices interpretated as one or more first storage devices)
and the secondary storage includes one or more second storage devices of a second type. (Bedadala, [0096], “ one or more secondary storage devices 108 configured to create and store one or more secondary copies ”)
Regarding claim 7, the combination Bedadala and Mishra teaches the claimed invention as claimed in The method of Claim 1, further comprising: determining that a frequency of access of particular data stored in the secondary storage is greater than or equal to a threshold frequency of access (Bedadala,[0040], “Some data, however, may not be accessed as frequently as other data. Because some data may not be accessed as frequently, this data can be transferred away from the primary storage device” furthermore, “storage policies to determine which data to transfer from a primary storage device to a secondary storage device. For example, a client computing device can be configured with one or more storage policies. A storage policy generally comprises a data structure or other information source that defines (or includes information sufficient to determine) a set of preferences or other criteria for performing the data transfer.” Examiners Remarks: Wherein the storage policy determines that a frequency of access of particular data stored in the secondary storage and the associated threshold of the frequency of access); and
in response to determining that the frequency of access of the particular data is greater than or equal to the threshold frequency of access (Bedadala, [0040], “storage policies to determine which data to transfer from a primary storage device to a secondary storage device. For example, a client computing device can be configured with one or more storage policies. A storage policy generally comprises a data structure or other information source that defines (or includes information sufficient to determine) a set of preferences or other criteria for performing the data transfer.”), migrating the particular data from the secondary storage to the primary storage. (Bedadala, [0163], “ data movement operations can include operations in which stored data is copied, migrated, or otherwise transferred from one or more first storage devices to one or more second storage devices, such as from primary storage device(s) 104 to secondary storage device(s) 108, from secondary storage devices 108 to primary storage devices 104”)
Regarding claim 8, the combination Bedadala and Mishra teaches the claimed invention as claimed in The method of Claim 1, wherein in response to determining that the received data is to be used for fine-tuning the LLM, the method further comprises: (Mishra, [0004], “ it can be assumed that each subsequent response is an improvement on the preceding response. Some of these implementations described herein relate to using the described techniques at inference time, to provide an improved response to a user that is responsive to the user's NL based input”, Furthermore, “the described techniques to generate training data by storing the NL based input along with an improved response that is responsive to the NL based input as an instance of training data. This training data can be used to fine-tune an LLM,” Examiner Remarks: wherein in the determination for fine-tuning the LLM is the subsequent NL based input, and if there is no subsequent NL input there is not fine-tuning of the LLM )
temporarily storing the received data in an intermediate storage of the storage system for fine-tuning the LLM (Bedadala, [0093], “Primary storage devices 104 storing primary data 112 may be relatively fast and/or expensive technology (e.g., flash storage, a disk drive, a hard-disk storage array, solid state memory, etc.), typically to support high-performance live production environments. Primary data 112 may be highly changeable and/or may be intended for relatively short term retention” Examiners Remarks: Under BRI the primary storage also act as an intermediate storage for a later utilization of data to fine-tine the LLM. ) ;
and storing the received data in the secondary storage after fine-tuning the LLM using the received data. (Bedadala,[0044], “use the trained data storage machine learning model [the LLM] in place of the storage polic(ies) to determine which data to store in a secondary storage”
Examiners Remarks: wherein the received data is used to train the LLM and that data is determined to be stored in a secondary storage)
Regarding claim 9, the combination Bedadala and Mishra teaches the claimed invention as claimed in The method of Claim 8, further comprising providing a batch of accumulated data stored in the intermediate storage to at least one processor for fine- tuning the LLM, the batch of accumulated data including the temporarily stored received data. ( Bedadala, [0093], “ Primary storage [the intermediate storage] devices 104 storing primary data Primary data [received data and a batch of accumulated data ] 112 may be highly changeable and/or may be intended for relatively short term retention (e.g., hours, days, or weeks).”) Furthermore (Bedadala, [0094], “primary storage device[at least one processor] 104 is dedicated to an associated client computing device 102” Examiner Remarks: Wherein the primary storage device acts an intermediate storage for a batch of accumulated data and received data. Furthermore, the primary storage device is associated with at least one processor)
Regarding claim 10, Bedadala teaches A system, comprising: a primary storage configured to store data (Bedadala,[0087], “ store the generated data to primary storage” Furthermore, Bedadala, [FIG. 1D], Examiners Remarks: displays a system including the, a primary and secondary storage
PNG
media_image2.png
604
587
media_image2.png
Greyscale
); and
a secondary storage configured to store data that is less frequently accessed than data stored in the primary storage (Bedadala,[0040], “Some data, however, may not be accessed as frequently as other data. Because some data may not be accessed as frequently, this data can be transferred away from the primary storage device”, Furthermore, “typical information management system can use one or more storage policies to determine which data to transfer from a primary storage device to a secondary storage device.”) ; and
at least one processor, individually or in combination, configured to (Bedadala,[0081], “Any given computing device comprises one or more processors”):
receive data from an application to store the received data (Bedadala, [0087], “a client may be said to generate data and to store the generated data”);
store the received data in the secondary storage. (Bedadala,[0319], “data objects to store in a secondary storage device”.)
Bedadala does not expressly teach determine whether to use the received data to fine-tune a Large Language Model (LLM); and in response to determining that the received data is to be used to fine- tune the LLM,
However, Mishra teaches, determine whether to use the received data to fine-tune a Large Language Model (LLM) ) (Mishra, [0004], “ it can be assumed that each subsequent response is an improvement on the preceding response. Some of these implementations described herein relate to using the described techniques at inference time, to provide an improved response to a user that is responsive to the user's NL based input”, Furthermore, “the described techniques to generate training data by storing the NL based input along with an improved response that is responsive to the NL based input as an instance of training data. This training data can be used to fine-tune an LLM,” wherein in the determination for fine-tuning the LLM is the subsequent NL based input, and if there is no subsequent NL input there is not fine-tuning of the LLM ); and
in response to determining that the received data is to be used to fine- tune the LLM, (Mishra, [0004], “ it can be assumed that each subsequent response is an improvement on the preceding response. Some of these implementations described herein relate to using the described techniques at inference time, to provide an improved response to a user that is responsive to the user's NL based input”, Furthermore, “the described techniques to generate training data by storing the NL based input along with an improved response that is responsive to the NL based input as an instance of training data. This training data can be used to fine-tune an LLM,” Examiner Remarks: wherein in the determination for fine-tuning the LLM is the subsequent NL based input, and if there is no subsequent NL input there is not fine-tuning of the LLM ).
Because Bedadala and Mishra are analogous art and within the same field of endeavor, specifically computer-implemented systems for the efficient utilization of large language models in storage systems. Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine Bedadala’s, machine learning based on data storage with Mishra’s, efficient fine-tuning of large language models (LLM). This modification would have been motivated by the desire to train a large language model (LLM) for performing specific task and reducing usage of computational resources. (Mishra, [0002].)
Regarding claims 11 - 13, and 15 - 20, these claims are system claims that corresponds to the methods of claims 2 – 4, and 6 - 9 , respectively. Therefore, claims 11 - 13, and 15 - 20, are rejected for the same reason as claims 2 – 4, and 6 - 9 above.
Claims 5 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Bedadala et al. (US 20190250839 A1) (hereinafter Bedadala), in view of Mishra et al. (US 20250045534 A1) (hereinafter Mishra) as applied to claim 1 and 10 above, and in further view of Waasdorp (US 20140129816 A1) (hereinafter Waasdorp).
Regarding claim 5, the combination Bedadala and Mishra teaches the claimed invention as claimed in the method of Claim 1, wherein the secondary storage is a first partition of a storage device of the storage system and the primary storage is a second partition of the storage device (Bedadala,[0293], “. Each serve [the storage system] may be configured for reading the at least one unmodified portion [second partition] of the operating system installation image from the primary storage drive, wherein each server is further configured for reading the at least one modified portion [first partition] of the operating system installation image from the secondary storage drive.”),
The combination Bedadala and Mishra teaches does not teach and wherein the secondary storage provides a higher data storage density than the primary storage.
However, Waasdorp teaches and wherein the secondary storage provides a higher data storage density than the primary storage (Waasdorp, [0013], “the secondary storage drive 104 includes at least ten times as much memory as the primary storage drive 102”)
Because Bedadala, Mishra and Waasdorp are analogous art and within the same field of endeavor, specifically computer-implemented systems for storage utilization. Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the dense data capacity management in Waasdorp’s Multi- server system to, Bedadala and Mishra‘ s, efficient utilization of large language models in storage systems. This modification would have been motivated by the desire to increase system performance by reducing latencies. (Waasdorp, [0002])
Regarding claim 14, this claim is system claim that corresponds to the methods of claim 5, respectively. Therefore, claim 14 is rejected for the same reason as claim 5 above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Chen, (US 20250077792 A1) discloses technologies capable of a training pipeline to fine-tune a machine learning model given a limited set of domain-specific data.
Posner, (US 20230148271 A1) discloses techniques for improving system performance based on data characteristics.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHAWN BLAIN whose telephone number is (571)270-1815. The examiner can normally be reached Mon - Fri: 8AM - 5PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Jennifer Welch can be reached at (571) 272-7212. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/S.B./Examiner, Art Unit 2143
/JENNIFER N WELCH/Supervisory Patent Examiner, Art Unit 2143
1 Claim language that is NOT Bolded and are in brackets i.e. {…}, are NOT taught by Bedadala
2 Claim language that is NOT Bold and are in brackets i.e. {…}, are NOT taught by Bedadala
