Artificial intelligence model and data collection/development platform
In some embodiments, a service platform that facilitates artificial intelligence model and data collection and collection may be provided. Input/output information derived from machine learning models may be obtained via the service platform. The input/output information may indicate (i) first items provided as input to at least one model of the machine learning models, (ii) first prediction outputs derived from the at least one model's processing of the first items, (iii) second items provided as input to at least another model of the machine learning models, (iv) second prediction outputs derived from the at least one other model's processing of the second items, and (v) other inputs and outputs. The input/output information may be provided via the service platform to update a first machine learning model. The first machine learning model may be updated based on the input/output information being provided as input to the first machine learning model.
1. A computing system comprising:
a storage configured to store a collection of training data that is previously consumed by machine learning models during training of the machine learning models via a host platform; and
a processor configured to;
receive a request to train a first machine learning model, the request comprising one or more parameters of input data to be used to train the first machine learning model,
select a training data set previously consumed by a second machine learning model based on input items within the training data set matching the one or more parameters of the input data in the received request,
train the first machine learning model via execution of the first machine learning model on the selected training data set via the host platform;
automatically test, in response to the training of the first machine learning model on the selected training data set, the trained first machine learning model to determine a predictive accuracy of the trained first machine learning model;
determine, based on the determined predictive accuracy of the trained first machine learning model and a predictive accuracy of the second machine learning model, that the trained first machine learning model is more accurate than the second machine learning model;
merge, automatically in response to the determination that the trained first machine learning model is more accurate than the second machine learning model, components from the first machine learning model with the second machine learning model to obtain an updated instance of the second machine learning model;
replace, in a memory storing the second machine learning model, the second machine learning model with the updated instance of the second machine learning model, the updated instance of the second machine learning model including a distinct version indicator associated therewith;
present a user-selectable model representation of the updated instance of the second machine learning model via a user interface;
receive, via the user interface, an indication of a selection of the updated instance of the second machine learning model; and
generate, in response to the reception of the selection of the updated instance of the second machine learning model, at least a portion of a software application.
2. The computing system of claim 1 , wherein the first machine learning model is a separate instance of a same model type as the second machine learning model.
3. The computing system of claim 1 , wherein the processor is further configured to remove a subset of data of the selected training data set prior to execution of the first machine learning model, and train the first machine learning model on remaining data within the selected training data set.
4. The computing system of claim 3 , wherein the processor is further configured to execute the trained first machine learning model on the removed subset of data to determine an accuracy of the trained first machine learning model.
5. The computing system of claim 1 , wherein the processor is configured to select the training data set via a machine learning model executing on the host platform.
6. The computing system of claim 1 , wherein the processor is configured to receive the request to train the first machine learning model from a search input via a search tool on a website.
7. A method comprising:
storing a collection of training data that is previously consumed by machine learning models during training of the machine learning models via a host platform;
receiving a request to train a first machine learning model, the request comprising one or more parameters of input data to be used to train the first machine learning model;
selecting a training data set previously consumed by a second machine learning model based on input items within the training data set matching the one or more parameters of the input data in the received request; and
training the first machine learning model via execution of the first machine learning model on the selected training data set via the host platform;
automatically testing, in response to the training of the first machine learning model on the selected training data set, the trained first machine learning model to determine a predictive accuracy of the trained first machine learning model;
determining, based on the determined predictive accuracy of the trained first machine learning model and a predictive accuracy of the second machine learning model, that the trained first machine learning model is more accurate than the second machine learning model;
merging, automatically in response to the determination that the trained first machine learning model is more accurate than the second machine learning model, components from the first machine learning model with the second machine learning model to obtain an updated instance of the second machine learning model;
replacing, in a memory storing the second machine learning model, the second machine learning model with the updated instance of the second machine learning model, the updated instance of the second machine learning model including a distinct version indicator associated therewith;
presenting a user-selectable model representation of the updated instance of the second machine learning model via a user interface;
receiving, via the user interface, an indication of a selection of the updated instance of the second machine learning model; and
generating, in response to the reception of the selection of the updated instance of the second machine learning model, at least a portion of a software application.
8. The method of claim 7 , wherein the first machine learning model is a separate instance of a same model type as the second machine learning model.
9. The method of claim 7 , wherein the training comprises removing a subset of data of the selected training data set prior to execution of the first machine learning model,
and training the first machine learning model on remaining data within the selected training data set.
10. The method of claim 9 , wherein the method further comprises executing the trained first machine learning model on the removed subset of data to determine an accuracy of the trained first machine learning model.
11. The method of claim 7 , wherein the selecting is performed by a third machine learning model executing on the host platform.
12. The method of claim 7 , wherein the receiving the request to train the first machine learning model comprises receiving a search input via a search tool on a website.
13. A non-transitory storage media comprising instructions which when executed by a processor cause a computer to perform a method comprising:
storing a collection of training data that is previously consumed by machine learning models during training of the machine learning models via a host platform;
receiving a request to train a first machine learning model, the request comprising one or more parameters of input data to be used to train the first machine learning model;
selecting a training data set previously consumed by a second machine learning model based on input items within the training data set matching the one or more parameters of the input data in the received request; and
training the first machine learning model via execution of the first machine learning model on the selected training data set via the host platform;
automatically testing, in response to the training of the first machine learning model on the selected training data set, the trained first machine learning model to determine a predictive accuracy of the trained first machine learning model;
determining, based on the determined predictive accuracy of the trained first machine learning model and a predictive accuracy of the second machine learning model, that the trained first machine learning model is more accurate than the second machine learning model;
merging, automatically in response to the determination that the trained first machine learning model is more accurate than the second machine learning model, components from the first machine learning model with the second machine learning model to obtain an updated instance of the second machine learning model;
replacing, in a memory storing the second machine learning model, the second machine learning model with the updated instance of the second machine learning model, the updated instance of the second machine learning model including a distinct version indicator associated therewith;
presenting a user-selectable model representation of the updated instance of the second machine learning model via a user interface;
receiving, via the user interface, an indication of a selection of the updated instance of the second machine learning model; and
generating, in response to the reception of the selection of the updated instance of the second machine learning model, at least a portion of a software application.
14. The non-transitory storage media of claim 13 , wherein the first machine learning model is a separate instance of a same model type as the second machine learning model.