Hands-on artificial intelligence education service

Abstract

Indications of sample machine learning models which create synthetic content items are provided via programmatic interfaces. A representation of a synthetic content item produced by one of the sample models in response to input obtained from a client of a provider network is presented. In response to a request from the client, a machine learning model is trained to produce additional synthetic content items.

1 - 20 . (canceled) 
     
     
         21 . A computer-implemented method, comprising:
 training one or more generative machine learning models to produce, in response to input content, a corresponding output content, wherein the output content comprises a newly-generated portion of content which shares one or more properties with the input content and thereby extends the input content, and wherein the one or more generative machine learning models include a transformer;   obtaining a particular input content via one or more programmatic interfaces at a network-accessible service; and   presenting, via the one or more programmatic interfaces, a particular output content generated at the network-accessible service using one or more trained versions of the one or more generative machine learning models, wherein the particular output content is produced by the one or more trained versions in response to the particular input content.   
     
     
         22 . The computer-implemented method as recited in  claim 21 , wherein the particular output content is produced at least in part by the transformer, the computer-implemented method further comprising:
 providing, by the network-accessible service via the one or more programmatic interfaces, an indication of an alternative generative machine learning model which can be utilized to generate output content corresponding to the particular input content, wherein the alternative generative machine learning model does not include a transformer; and   presenting, by the network-accessible service via the one or more programmatic interfaces in response to a request to utilize the alternative generative machine learning model, additional output content corresponding to the particular input content, wherein the additional output content is generated using the alternative generative machine learning model.   
     
     
         23 . The computer-implemented method as recited in  claim 22 , wherein the alternative generative machine learning model comprises one or more of: (a) a generative adversarial network (b) a variational auto-encoder or (c) an autoregressive model. 
     
     
         24 . The computer-implemented method as recited in  claim 21 , further comprising:
 presenting, by the network-accessible service via the one or more programmatic interfaces, an indication of one or more hyper-parameters of the one or more generative machine learning models.   
     
     
         25 . The computer-implemented method as recited in  claim 21 , further comprising:
 presenting, by the network-accessible service via the one or more programmatic interfaces, an indication of a change in the particular output content resulting from a change to a hyper-parameter of the one or more generative machine learning models.   
     
     
         26 . The computer-implemented method as recited in  claim 21 , further comprising:
 presenting, by the network-accessible service via the one or more programmatic interfaces, an indication of a structure of the one or more generative machine learning models, including a number of layers of a neural network utilized in the one or more generative machine learning models.   
     
     
         27 . The computer-implemented method as recited in  claim 21 , wherein the particular input content comprises one or more of: (a) audio, (b) an image, (c) a drawing, (d) a painting or (e) text. 
     
     
         28 . A system, comprising:
 one or more computing devices, wherein the one or more computing devices include instructions that upon execution on or across the one or more computing devices:
 train one or more generative machine learning models to produce, in response to input content, a corresponding output content, wherein the output content comprises a newly-generated portion of content which shares one or more properties with the input content and thereby extends the input content, and wherein the one or more generative machine learning models include a transformer; 
 obtain a particular input content via one or more programmatic interfaces at a network-accessible service; and 
 present, via the one or more programmatic interfaces, a particular output content generated at the network-accessible service using one or more trained versions of the one or more generative machine learning models, wherein the particular output content is produced by the one or more trained versions in response to the particular input content. 
   
     
     
         29 . The system as recited in  claim 28 , wherein the particular output content is produced at least in part by the transformer, and wherein the one or more computing devices include further instructions that upon execution on or across the one or more computing devices:
 provide, by the network-accessible service via the one or more programmatic interfaces, an indication of an alternative generative machine learning model which can be utilized to generate output content corresponding to the particular input content, wherein the alternative generative machine learning model does not include a transformer; and   present, by the network-accessible service via the one or more programmatic interfaces in response to a request to utilize the alternative generative machine learning model, additional output content corresponding to the particular input content, wherein the additional output content is generated using the alternative generative machine learning model.   
     
     
         30 . The system as recited in  claim 29 , wherein the alternative generative machine learning model comprises one or more of: (a) a generative adversarial network (b) a variational auto-encoder or (c) an autoregressive model. 
     
     
         31 . The system as recited in  claim 28 , wherein the one or more computing devices include further instructions that upon execution on or across the one or more computing devices:
 present, by the network-accessible service via the one or more programmatic interfaces, an indication of one or more hyper-parameters of the one or more generative machine learning models.   
     
     
         32 . The system as recited in  claim 28 , wherein the one or more computing devices include further instructions that upon execution on or across the one or more computing devices:
 present, by the network-accessible service via the one or more programmatic interfaces, an indication of a change in the particular output content resulting from a change to a hyper-parameter of the one or more generative machine learning models.   
     
     
         33 . The system as recited in  claim 28 , wherein the one or more computing devices include further instructions that upon execution on or across the one or more computing devices:
 present, by the network-accessible service via the one or more programmatic interfaces, an indication of a structure of the one or more generative machine learning models, including a number of layers of a neural network utilized in the one or more generative machine learning models.   
     
     
         34 . The system as recited in  claim 28 , wherein the particular input content comprises one or more of: (a) audio, (b) an image, (c) a drawing, (d) a painting or (e) text. 
     
     
         35 . One or more non-transitory computer-accessible storage media storing program instructions that when executed on or across one or more processors:
 train one or more generative machine learning models to produce, in response to input content, a corresponding output content, wherein the output content comprises a newly-generated portion of content which shares one or more properties with the input content and thereby extends the input content, and wherein the one or more generative machine learning models include a transformer;   obtain a particular input content via one or more programmatic interfaces at a network-accessible service; and   present, via the one or more programmatic interfaces, a particular output content generated at the network-accessible service using one or more trained versions of the one or more generative machine learning models, wherein the particular output content is produced by the one or more trained versions in response to the particular input content.   
     
     
         36 . The one or more non-transitory computer-accessible storage media as recited in  claim 35 , wherein the particular output content is produced at least in part by the transformer, and wherein the one or more non-transitory computer-accessible storage media store further program instructions that when executed on or across the one or more processors:
 provide, by the network-accessible service via the one or more programmatic interfaces, an indication of an alternative generative machine learning model which can be utilized to generate output content corresponding to the particular input content, wherein the alternative generative machine learning model does not include a transformer; and   present, by the network-accessible service via the one or more programmatic interfaces in response to a request to utilize the alternative generative machine learning model, additional output content corresponding to the particular input content, wherein the additional output content is generated using the alternative generative machine learning model.   
     
     
         37 . The one or more non-transitory computer-accessible storage media as recited in  claim 36 , wherein the alternative generative machine learning model comprises one or more of: (a) a generative adversarial network (b) a variational auto-encoder or (c) an autoregressive model. 
     
     
         38 . The one or more non-transitory computer-accessible storage media as recited in  claim 35 , storing further program instructions that when executed on or across the one or more processors:
 present, by the network-accessible service via the one or more programmatic interfaces, an indication of a change in the particular output content resulting from a change to a hyper-parameter of the one or more generative machine learning models.   
     
     
         39 . The one or more non-transitory computer-accessible storage media as recited in  claim 35 , storing further program instructions that when executed on or across the one or more processors:
 present, by the network-accessible service via the one or more programmatic interfaces, an indication of a structure of the one or more generative machine learning models, including a number of layers of a neural network utilized in the one or more generative machine learning models.   
     
     
         40 . The one or more non-transitory computer-accessible storage media as recited in  claim 35 , wherein the particular input content comprises one or more of: (a) audio, (b) an image, (c) a drawing, (d) a painting or (e) text.