3results about How to "Improve labeling ability" patented technology

The invention discloses an intelligent labeling method and system based on multiple modes, and relates to the technical field of data labeling, and the method comprises the steps: collecting multi-mode data from multiple sources, preprocessing the collected data, labeling the multi-mode data according to task requirements, and dividing a labeled data set into a training set, a verification set and a test set; selecting a model architecture, setting training parameters, training the model by using training set data, and optimizing the model parameters; regularly evaluating model performance through a verification set, and adjusting a training strategy; evaluating the performance of the model on the test set, and improving the model according to a test result. According to the method, multi-modal large model fine adjustment is applied to labeling, an existing manual labeling mode is replaced, the labor cost of an enterprise is greatly saved, intelligent labeling can be used for coarse screening before data preparation, secondary verification can also be conducted on a small model result, the generalization ability of the model is improved, and a better labeling effect is brought.

Embodiments of the present specification provide a task processing model training method, a task processing method and a data labeling model training method. The task processing model training method comprises: obtaining training data of a target training task; inputting the training data into a data labeling model to obtain a training labeling result of the training data, wherein the data labeling model is trained based on sample data, sample auxiliary data of the sample data and a sample labeling result, the sample auxiliary data is obtained through multi-round argumentation of multiple argumentation roles based on the sample data, and the argumentation roles have different argumentation viewpoints; and training an initial processing model according to the training data and the training labeling result to obtain a task processing model. The labeling capability of the data labeling model and the robustness and generalization capability of the data labeling model when facing new data are improved through the multi-round argumentation process, so that the data labeling model can generate reliable training labeling results consistent with human intentions, and the accuracy of the task processing model is improved.