Method and apparatus for updating deep learning model

1. A method for updating a deep learning model, comprising:

executing following updating: acquiring a training dataset under a preset path, training a preset deep learning model based on the training dataset to obtain a new deep learning model; updating the preset deep learning model to the new deep learning model; increasing a number of training by 1; determining whether the number of training reaches a threshold of the number of training; stopping executing the updating if the number of training reaches the threshold, wherein an initial number of the number of training is zero; and

if the number of training fails to reach the threshold, continuing to execute the updating after a preset time interval, such that, before the updating is stopped, the updating is performed periodically by using the preset time interval as a period,

wherein when the new deep learning model is obtained, the updating further comprises:

saving the new deep learning model; and

after updating the preset deep learning model to the new deep learning model, the updating further comprises:

sending, to a user the preset deep learning model attributed to, a model updating prompt message comprising a save path, a save title, and a timestamp of the employed training dataset of the new deep learning model.

2. The method according to claim 1 , wherein the training dataset under the preset path is provided with a dataset identifier containing a time stamp; and

the acquiring a training dataset under a preset path comprises:

acquiring a training dataset with the time stamp contained in the dataset identifier matching a current time stamp under the preset path.

3. The method according to claim 2 , wherein the acquiring a training dataset with the time stamp contained in the dataset identifier matching a current time stamp under the preset path comprises:

calculating a similarity between the current time stamp and the time stamp contained in a dataset identifier of each of a plurality of training datasets under the preset path, and determining a time stamp with a highest similarity to the current time stamp as a target time stamp matching the current time stamp; and

acquiring a training dataset with a dataset identifier containing the target time stamp under the preset path.

4. The method according to claim 1 , wherein the preset deep learning model has a training script corresponding to the preset deep learning model; and

the training a preset deep learning model based on the training dataset to obtain a new deep learning model comprises:

running the training script to train the preset deep learning model based on the training dataset to obtain the new deep learning model.

5. The method according to claim 1 , wherein the training dataset under the preset path is generated through regularly running a preset training dataset generation code.

6. The method according to claim 2 , wherein the acquiring a training dataset with the time stamp contained in the dataset identifier matching a current time stamp under the preset path comprises:

acquiring a training dataset with the time stamp contained in the dataset identifier identical with the current time stamp under the preset path.

7. An apparatus for updating a deep learning model, comprising:

at least one processor; and

a memory storing instructions, the instructions when executed by the at least one processor, cause the at least one processor to perform operations, the operations comprising:

executing following updating: acquiring a training dataset under a preset path, training a preset deep learning model based on the training dataset to obtain a new deep learning model; updating the preset deep learning model to the new deep learning model; increasing a number of training by 1; determining whether the number of training reaches a threshold of the number of training; stopping executing the updating if the number of training reaches the threshold, wherein an initial number of the number of training is zero; and

if the number of training fails to reach the threshold, continuing to execute the updating after a preset time interval, such that, before the updating is stopped, the updating is performed periodically by using the preset time interval as a period,

wherein when the new deep learning model is obtained, the updating further comprises:

saving the new deep learning model; and

after updating the preset deep learning model to the new deep learning model, the updating further comprises:

sending, to a user the preset deep learning model attributed to, a model updating prompt message comprising a save path, a save title, and a timestamp of the employed training dataset of the new deep learning model.

8. The apparatus according to claim 7 , wherein the training dataset under the preset path is provided with a dataset identifier containing a time stamp; and

the acquiring a training dataset under a preset path comprises:

acquiring a training dataset with the time stamp contained in the dataset identifier matching a current time stamp under the preset path.

9. The apparatus according to claim 8 , wherein the acquiring a training dataset with the time stamp contained in the dataset identifier matching a current time stamp under the preset path comprises:

calculating a similarity between the current time stamp and the time stamp contained in a dataset identifier of each of a plurality of training datasets under the preset path, and determining a time stamp with a highest similarity to the current time stamp as a target time stamp matching the current time stamp; and

acquiring a training dataset with a dataset identifier containing the target time stamp under the preset path.

10. The apparatus according to claim 7 , wherein the preset deep learning model has a training script corresponding to the preset deep learning model; and

the training a preset deep learning model based on the training dataset to obtain a new deep learning model comprises:

running the training script to train the preset deep learning model based on the training dataset to obtain the new deep learning model.

11. The apparatus according to claim 7 , wherein the training dataset under the preset path is generated through regularly running a preset training dataset generation code.

12. A non-transitory computer readable storage medium storing a computer program, wherein the computer program, when executed by a processor, cause the processor to perform operations, the operation comprising:

executing following updating: acquiring a training dataset under a preset path, training a preset deep learning model based on the training dataset to obtain a new deep learning model; updating the preset deep learning model to the new deep learning model; increasing a number of training by 1; determining whether the number of training reaches a threshold of the number of training; stopping executing the updating if the number of training reaches the threshold, wherein an initial number of the number of training is zero; and

if the number of training fails to reach the threshold, continuing to execute the updating after a preset time interval, such that, before the updating is stopped, the updating is performed periodically by using the preset time interval as a period,

wherein when the new deep learning model is obtained, the updating further comprises:

saving the new deep learning model; and

after updating the preset deep learning model to the new deep learning model, the updating further comprises:

sending, to a user the preset deep learning model attributed to, a model updating prompt message comprising a save path, a save title, and a timestamp of the employed training dataset of the new deep learning model.