Acer

The disclosure provides a method for training an image generation model and an electronic apparatus (100), and the method includes the following steps. A plurality of first layer outputs of a first teacher model (TM1) and a plurality of second layer outputs of a second teacher model (TM2) are obtained by inputting a training image to the first teacher model (TM1) and the second teacher model (TM2). The training image is input to a student model (SM1) to obtain a plurality of third layer outputs of the student model (SM1). A first loss between the first layer outputs and the third layer outputs is determined according to a plurality of first loss weights. A second loss between the second layer outputs and the third layer outputs is determined. The student model (SM1) is updated by using the first loss and the second loss to train the student model (SM1).

Provided are a method for dynamically adjusting hardware configuration instruction and an electronic apparatus (100). The method for dynamically adjusting hardware configuration instructions is adapted to the electronic apparatus including a basic input-output system (BIOS) device (110) and a hardware device (120). The method includes the following steps. A binary raw data file (F1) is generated according to a hardware configuration instruction set of the hardware device, wherein the hardware configuration instruction set includes at least one hardware configuration instruction. The binary raw data file is written to BIOS variable storage space (31) of the BIOS device. By executing BIOS code (B1), the hardware configuration instruction in the binary raw data file is read from the BIOS variable storage space. By executing the BIOS code, the hardware configuration instruction read from the BIOS variable storage space is written into a register (R1) of the hardware device.

A user equipment (UE) receives physical downlink (DL) shared channel (PDSCH) in a DL slot as scheduled using a DL control indicator (DCI). The UE identifies that the DL slot is part of a group of DL slots based on the DCI and/or another DCI corresponding to the group. The UE identifies one or more uplink (UL) slots at which to transmit hybrid automatic repeat request (HARQ) feedback corresponding to the slots in the group, and transmits the HARQ feedback information at one of these UL slots. In some cases, if HARQ feedback information is not received, a semi-static codebook may prevent HARQ payload size misalignment.

A wireless communication method performed by a communication device (16, 20) includes performing at least one communication operation with a network (14, 20) via a first channel, CH (1002); and operating on a second CH according to at least one of an indication, a temporal constraint, and an interference condition (1004); wherein at least one non-primary CH access, NPCA, CH is obtained by the network.

A wireless communication method performed by a communication device (16, 20) includes performing at least one communication operation with a network (14, 20) via a first channel, CH (1002); and operating on a second CH according to at least one of an indication, a temporal constraint, and an interference condition (1004).

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …