Abstract: This application provides a method for training a semantic understanding model, including: obtaining a first training sample set; performing denoising processing on the first training sample set to form a corresponding second training sample set; processing the second training sample set by using a semantic understanding model, to determine initial parameters of the semantic understanding model; processing the second training sample set by using the semantic understanding model in response to the initial parameters of the semantic understanding model, to determine update parameters of the semantic understanding model; and iteratively updating a semantic representation layer network parameter and a task-related output layer network parameter of the semantic understanding model by using the second training sample set and according to the update parameters of the semantic understanding model.

Abstract: The present disclosure discloses a W-containing high-nickel ternary cathode material, including both spherical secondary particles and single-crystal particles. There is basically no W inside the single-crystal particles, and the spherical secondary particles are doped with W.

Abstract: The present invention provides use of a multi-target protein kinase inhibitor compound A or a pharmaceutically acceptable salt thereof in the manufacture of a medicament for treating leukemia, and provides a method for treating acute myeloid leukemia, especially acute myeloid leukemia with FLT3 mutation, with compound A. In clinical trials, compound A has certain efficacies both on acute myeloid leukemia with FLT3-ITD mutation and/or FLT3-TKD mutation, and on DEK-CAN positive acute myeloid leukemia with FLT3-ITD mutation. Patients with relapsed and/or refractory acute myeloid leukemia who have failed treatment previously with Type II FLT3 inhibitors (e.g., sorafenib) can still clinically benefit from the treatment with compound A.

Abstract: This application provides a method for training a semantic understanding model, including: obtaining a first training sample set; performing denoising processing on the first training sample set to form a corresponding second training sample set; processing the second training sample set by using a semantic understanding model, to determine initial parameters of the semantic understanding model; processing the second training sample set by using the semantic understanding model in response to the initial parameters of the semantic understanding model, to determine update parameters of the semantic understanding model; and iteratively updating a semantic representation layer network parameter and a task-related output layer network parameter of the semantic understanding model by using the second training sample set and according to the update parameters of the semantic understanding model.