Abstract: State of art approaches independently use a Pruning-weight Clustering-Quantization (PCQ) or Knowledge Distillation (KD) for model optimization and require critical manual intervention. Embodiments of the present disclosure provide a method and system for the two-step hierarchical model optimization approach for generating optimized model DL model. The method comprises a AutoPCQ technique followed by conditional application of an automated KD (AKD) technique. The AutoPCQ technique formulates a problem of configuration selection of the DL model as an optimization problem by iteratively applying Bayesian optimization and Reinforcement Learning. Further, the AKD technique formulates automated search of a student model as the optimization problem with the DL model representing a teacher model. A search space for the student model is defined by a restricted Neural Network Architecture Search that restricts the search space.

Abstract: Human-understandable explanations of Artificial Intelligence (AI) based models are crucial to building transparency and trust in AI based solutions. More importantly, these explanations need to be contextual, applicable to the domain the model is used in and relevant to the concerned stakeholder. Conventionally, there is a lack of communicating these explanations to various stakeholders in a language that they can understand and relate to. The present disclosure facilitates the conversational agents (chat bots) with intelligence and actions that would help them communicate the right information to the right stakeholder in the right way. In the present disclosure, contextual explanation for user queries is generated based on the output from AI models. Here, the impacting features are obtained from the explainer model associated with the prediction model and the contextual information is generated. Further, the contextual information is converted to the contextual explanation to the user.

Abstract: Business rules are currently not documented and are present only as knowledge with subject matter experts (SMEs). The knowledge can be lost with time if it is not extracted or recorded. Existing techniques are unable to extract tacit knowledge and to retain the domain flavor in extracted information. Present disclosure provides a method and a system for extracting tacit knowledge from historical data. The system represents each point in historical data as a large dimensional hyperspace which contains all unstructured information where tacit knowledge can exist. Then, system maps large dimensional hyperspace to smaller dimensional hyperspace using pre-trained large language model (LLM). Thereafter, system, based on the series of downstream tasks, generates a feedback loop to optimally compute dimension of the smaller dimensional hyperspace.

Abstract: This disclosure relates generally to system and method to track performance of an AI application during development and production lifecycle. Libraries and tools that are used to develop an enterprise AI application are rapidly expanding across vendors and open-source community. These libraries are coming up with new features and breaking changes with new versions making it difficult to application developer and enterprise runtime executor. The method of the present disclosure receives an enterprise artificial intelligence (AI) application and corresponding software environment as input generate a logger plug-in using one or more fine-tuned large language models (LLMs) based on the plurality of instructions provided by the one or more instructors. Additionally, the logger plug-in is utilized to validate correctness of errors and then the logger plug-in is executed with the enterprise AI application to log and track performance of the enterprise AI application.

Abstract: This disclosure relates generally to method and system to generate persona-based commentary data for machine learning model card document. Existing techniques on model card are designed mainly for personas and understanding section of the model card document requires a certain level of expertise in machine learning. The method of the present disclosure receives a model card document comprising a plurality of sections and the model card document corresponds to a persona. Each section of the model card document obtains a metadata for the persona. The data curator machine learning model automatically generates a persona-based report trajectory for a plurality of sections of the metadata and a plurality of schema rules to generate a prompt template. The commentary generator ML model generates one or more commentary data for each section associated with the prompt template corresponding to the persona.

Abstract: This disclosure relates generally relates to method and system to process asynchronous and distributed training tasks. Training a large-scale deep neural network (DNN) model with large-scale training data is time-consuming. The method creates a work queue (Q) with a set of predefined number of tasks comprising a training data. Here, set of central processing units (CPUs) information and a set of graphics processing units (GPUs) information are fetched from the current environment to initiate a parallel process asynchronously on the work queue (Q) to train a set of deep learning models with optimized resources using a data pre-processing technique, to compute a transformed training data and training by using an asynchronous model training technique, the set of deep learning models on each GPU asynchronously with the transformed training data based on a set of asynchronous model parameters.

Abstract: Human-understandable explanations of Artificial Intelligence (AI) based models are crucial to building transparency and trust in AI based solutions. More importantly, these explanations need to be contextual, applicable to the domain the model is used in and relevant to the concerned stakeholder. Conventionally, there is a lack of communicating these explanations to various stakeholders in a language that they can understand and relate to. The present disclosure facilitates the conversational agents (chat bots) with intelligence and actions that would help them communicate the right information to the right stakeholder in the right way. In the present disclosure, contextual explanation for user queries is generated based on the output from AI models. Here, the impacting features are obtained from the explainer model associated with the prediction model and the contextual information is generated. Further, the contextual information is converted to the contextual explanation to the user.

Abstract: This disclosure relates generally relates to method and system to process asynchronous and distributed training tasks. Training a large-scale deep neural network (DNN) model with large-scale training data is time-consuming. The method creates a work queue (Q) with a set of predefined number of tasks comprising a training data. Here, set of central processing units (CPUs) information and a set of graphics processing units (GPUs) information are fetched from the current environment to initiate a parallel process asynchronously on the work queue (Q) to train a set of deep learning models with optimized resources using a data pre-processing technique, to compute a transformed training data and training by using an asynchronous model training technique, the set of deep learning models on each GPU asynchronously with the transformed training data based on a set of asynchronous model parameters.

Abstract: State of art approaches independently use a Pruning-weight Clustering-Quantization (PCQ) or Knowledge Distillation (KD) for model optimization and require critical manual intervention. Embodiments of the present disclosure provide a method and system for the two-step hierarchical model optimization approach for generating optimized model DL model. The method comprises a AutoPCQ technique followed by conditional application of an automated KD (AKD) technique. The AutoPCQ technique formulates a problem of configuration selection of the DL model as an optimization problem by iteratively applying Bayesian optimization and Reinforcement Learning. Further, the AKD technique formulates automated search of a student model as the optimization problem with the DL model representing a teacher model. A search space for the student model is defined by a restricted Neural Network Architecture Search that restricts the search space.