3 results about "Register based" patented technology

The application discloses a large language model weight compression method, an inference operation method, a system and a medium, and belongs to the technical field of large language model weight compression and inference operation optimization. The method comprises the following steps: acquiring a weight matrix of a large language model and constructing an exponential high-frequency window and a benchmark value, classifying and encoding each weight data into a fixed-length code word and splitting the fixed-length code word into an independent bitmap, synchronously generating a compact value stream and a rollback value stream, and packing the compact value stream and the rollback value stream into weight compression data; loading the weight compression data, generating a state mask and a channel prefix mask in a register based on the bitmap, calculating a reading offset by using a population count instruction, reading a compressed data segment from a corresponding data stream, reconstructing weight data in the register in combination with the code word, and finally directly inputting the weight data into a matrix multiplication unit for operation; therefore, by implementing the application, the problem of efficiency reduction caused by control flow divergence and redundant memory access in GPU inference of existing variable-length encoding can be solved, and the model inference efficiency under lossless compression can be improved.

The system according to this embodiment aims to provide quick and accurate answers based on various information sources. [Solution] The system according to the embodiment comprises a registration unit, an analysis unit, a generation unit, and a provision unit. The registration unit registers information sources. The analysis unit analyzes questions based on the information sources registered by the registration unit. The generation unit generates answers based on the questions analyzed by the analysis unit. The provision unit provides the answers generated by the generation unit.

The invention belongs to the technical field of outpatient registration and registration management, and discloses an intelligent hospital outpatient registration and registration information management system. The system comprises a patient doctor-seeing registration module, a department registration confirmation module, a patient identity classification module, a normal patient sorting module, a special patient sorting module, a re-visit patient adjustment module, a doctor-seeing sorting display module, a number calling mode analysis module and a data storage library. According to the method, whether the patient is allowed to be registered or not is judged based on the habitual diagnosis duration, the diagnosis time of the patient can be estimated more accurately through the analysis mode, the situation that resources are wasted or the patient waits for a long time due to a fixed number of registration is avoided, meanwhile, the registered patient is classified, and the registration efficiency is improved. According to the method, different modes of number calling sorting are carried out for different types of patients, number calling is carried out in an interleaving number calling mode, the analysis mode enables the special patients to be processed preferentially, and the overall efficiency of medical services is improved.