Class BaseModel
- Namespace
- GenerativeAI
- Assembly
- GenerativeAI.dll
The BaseModel class is an abstract representation of the model interactions in the GenerativeAI library. It serves as a foundational implementation that enables derived classes to interact with the GenerativeAI platform. This class inherits from BaseClient. It provides protected methods for use by derived classes, allowing them to handle content generation, token management, and content embedding functionalities. BaseModel is designed for flexible extension, enabling the development of specific AI model implementations.
public abstract class BaseModel : BaseClient
- Inheritance
-
BaseModel
- Derived
- Inherited Members
- Extension Methods
Constructors
BaseModel(IPlatformAdapter, HttpClient?, ILogger?)
Abstract base class for models that interact with generative AI functionalities, providing common implementation details for generating content, embedding content, and token counting operations.
protected BaseModel(IPlatformAdapter platform, HttpClient? httpClient, ILogger? logger = null)
Parameters
platformIPlatformAdapterhttpClientHttpClientloggerILogger
Remarks
This class extends BaseClient and serves as a foundational class for specific models like embedding or generative models. It provides core functionality for handling requests to the generative AI services, including streaming and non-streaming content generation, token counting, and content embedding.
Methods
BatchEmbedContentAsync(string, BatchEmbedContentRequest, CancellationToken)
Embeds a batch of content using the specified Generative AI model.
protected virtual Task<BatchEmbedContentsResponse> BatchEmbedContentAsync(string model, BatchEmbedContentRequest request, CancellationToken cancellationToken = default)
Parameters
modelstringThe name of the Generative AI Model to use for embedding the content. Format: "models/{model}".
requestBatchEmbedContentRequestThe batch BatchEmbedContentRequest containing the content and optional settings for embedding.
cancellationTokenCancellationTokenA token to monitor for cancellation requests.
Returns
- Task<BatchEmbedContentsResponse>
The BatchEmbedContentsResponse containing the content embeddings.
- See Also
CheckBlockedResponse(GenerateContentResponse?, string)
Checks the response from the generative AI model to determine if the response is blocked or invalid and throws an exception if so.
protected static void CheckBlockedResponse(GenerateContentResponse? response, string url)
Parameters
responseGenerateContentResponseThe GenerateContentResponse received from the generative AI model, which may contain content candidates.
urlstringThe URL of the request made to the generative AI model, used for error reporting.
Exceptions
- GenerativeAIException
Thrown if the response is blocked or invalid with details of the error.
CheckBlockedResponse(GenerateContentResponse?, Uri)
Checks if the response contains blocked content and handles accordingly.
protected void CheckBlockedResponse(GenerateContentResponse? response, Uri url)
Parameters
responseGenerateContentResponseThe response to check for blocked content.
urlUriThe URL of the request that generated the response.
CountTokensAsync(string, CountTokensRequest, CancellationToken)
Asynchronously counts the number of tokens in a given input using the specified model and request data.
protected virtual Task<CountTokensResponse> CountTokensAsync(string model, CountTokensRequest request, CancellationToken cancellationToken = default)
Parameters
modelstringThe name of the Generative AI model to use for token counting, in the format
models/{model}.requestCountTokensRequestThe CountTokensRequest containing the input data for token counting.
cancellationTokenCancellationTokenAn optional token to observe while waiting for the task to complete.
Returns
- Task<CountTokensResponse>
A CountTokensResponse containing details about the token count.
EmbedContentAsync(string, EmbedContentRequest, CancellationToken)
Embeds content using the specified Generative AI Model.
protected virtual Task<EmbedContentResponse> EmbedContentAsync(string model, EmbedContentRequest request, CancellationToken cancellationToken = default)
Parameters
modelstringThe name of the Generative AI Model to use for embedding the content. Format: "models/{model}".
requestEmbedContentRequestThe EmbedContentRequest containing the content and optional settings for embedding.
cancellationTokenCancellationTokenA token to monitor for cancellation requests.
Returns
- Task<EmbedContentResponse>
The EmbedContentResponse containing the generated text embedding vector.
- See Also
GenerateAnswerAsync(string, GenerateAnswerRequest, CancellationToken)
Generates a grounded answer from the model given an input GenerateAnswerRequest.
protected Task<GenerateAnswerResponse> GenerateAnswerAsync(string model, GenerateAnswerRequest request, CancellationToken cancellationToken = default)
Parameters
modelstringThe name of the Model to use for generating the grounded response. Format:
models/{model}.requestGenerateAnswerRequestThe GenerateAnswerRequest containing the input data.
cancellationTokenCancellationToken
Returns
- Task<GenerateAnswerResponse>
The GenerateAnswerResponse containing the model's answer.
- See Also
GenerateContentAsync(string, GenerateContentRequest, CancellationToken)
Generates a model response given an input GenerateContentRequest.
protected virtual Task<GenerateContentResponse> GenerateContentAsync(string model, GenerateContentRequest request, CancellationToken cancellationToken = default)
Parameters
modelstringThe name of the Generative AI Model to use for generating the completion. Format:
models/{model}.requestGenerateContentRequestThe GenerateContentRequest containing the content of the current conversation with the model.
cancellationTokenCancellationTokenAn optional token to observe while waiting for the task to complete.
Returns
- Task<GenerateContentResponse>
The GenerateContentResponse containing the model's response.
- See Also
GenerateContentStreamAsync(string, GenerateContentRequest, CancellationToken)
T Streams model responses chunk by chunk given an input GenerateContentRequest.
protected virtual IAsyncEnumerable<GenerateContentResponse> GenerateContentStreamAsync(string model, GenerateContentRequest request, CancellationToken cancellationToken = default)
Parameters
modelstringThe name of the Generative AI Model to use for generating the completion. Format: "models/{model}".
requestGenerateContentRequestThe GenerateContentRequest containing content of the current conversation with the model.
cancellationTokenCancellationTokenToken for cancelling the streaming process.
Returns
- IAsyncEnumerable<GenerateContentResponse>
An async stream of GenerateContentResponse items.
- See Also