mirror of
https://github.com/firebase/firebase-ios-sdk
synced 2026-11-11 22:55:59 +05:45
Andrew Heard
3796a55fdf
Implemented `GeminiLanguageModel`, providing an adapter for Google's Gemini models conforming to Apple's `FoundationModels.LanguageModel` protocol. This PR includes text to text generation capabilities; other modalities and Gemini features will be handled in follow-up PRs. **Note:** The `GeminiLanguageModel` implementation is guarded by the environment variable `GEMINI_LANGUAGE_MODEL`. By default none of the `GeminiLanguageModel` implementation is included. Key additions: - `GeminiLanguageModel`: Implements `LanguageModel` protocol with reasoning capabilities, configuring execution via `ModelResource` and `EndpointConfiguration`. - `GeminiLanguageModelExecutor`: Converts prompt instructions and history into Gemini streaming requests, yielding responses as `LanguageModelSession.ResponseStream`. - `GeminiTranscriptTranslator`: Translates Apple `Transcript` entries and segments into Gemini `Content` structures while preserving thought signatures and validating supported transcript features. - `GeminiLanguageModel.Error` and `GeminiErrorMapper`: Translates Gemini API, HTTP, and network errors into standard `LanguageModelError` cases (`rateLimited`, `timeout`, `guardrailViolation`, `refusal`, and `unsupportedLanguageOrLocale`) or `GeminiLanguageModel.Error` cases (`serviceUnavailable`, `networkFailure`, `modelNotFound`, `apiError`). The `APIError` payload is designed for seamless forward compatibility with both GenerateContent and the upcoming Interactions API. - Testing infrastructure: Adds `ConditionTrait.requireFoundationModels` to conditionally run Foundation Models tests only on supported platforms (macOS 27+, iOS 27+, watchOS 27+, visionOS 27+), along with shared test constants, mock fixtures, and comprehensive unit and integration test suites.
229 lines
9.1 KiB
Swift
229 lines
9.1 KiB
Swift
// Copyright 2026 Google LLC
|
|
//
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
// you may not use this file except in compliance with the License.
|
|
// You may obtain a copy of the License at
|
|
//
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
//
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
// See the License for the specific language governing permissions and
|
|
// limitations under the License.
|
|
|
|
import Foundation
|
|
|
|
/// An internal data model for `Candidate`.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Type: `GoogleAiGenerativelanguageV1betaCandidate`
|
|
///
|
|
/// A response candidate generated from the model.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Type: `GoogleCloudAiplatformV1beta1Candidate`
|
|
///
|
|
/// A response candidate generated from the model.
|
|
package struct Candidate: Codable, Sendable, Equatable, Hashable {
|
|
/// Output only. Index of the candidate in the list of response candidates.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Index of the candidate in the list of response candidates.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. The 0-based index of this candidate in the list of generated
|
|
/// responses. This is useful for distinguishing between multiple candidates
|
|
/// when `candidate_count` > 1.
|
|
package let index: Int?
|
|
|
|
/// Output only. Generated content returned from the model.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Generated content returned from the model.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. The content of the candidate.
|
|
package let content: Content?
|
|
|
|
/// Optional. Output only. The reason why the model stopped generating tokens.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Optional. Output only. The reason why the model stopped generating tokens.
|
|
///
|
|
/// If empty, the model has not stopped generating tokens.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. The reason why the model stopped generating tokens. If empty,
|
|
/// the model has not stopped generating.
|
|
package let finishReason: FinishReason?
|
|
|
|
/// Optional. Output only. Details the reason why the model stopped generating tokens.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Optional. Output only. Details the reason why the model stopped generating tokens.
|
|
/// This is populated only when `finish_reason` is set.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. Describes the reason the model stopped generating tokens in
|
|
/// more detail. This field is returned only when `finish_reason` is set.
|
|
package let finishMessage: String?
|
|
|
|
/// List of ratings for the safety of a response candidate.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// List of ratings for the safety of a response candidate.
|
|
///
|
|
/// There is at most one rating per category.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. A list of ratings for the safety of a response candidate.
|
|
///
|
|
/// There is at most one rating per category.
|
|
package let safetyRatings: [SafetyRating]?
|
|
|
|
/// Output only. Citation information for model-generated candidate.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Citation information for model-generated candidate.
|
|
///
|
|
/// This field may be populated with recitation information for any text
|
|
/// included in the `content`. These are passages that are "recited" from
|
|
/// copyrighted material in the foundational LLM's training data.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. A collection of citations that apply to the generated content.
|
|
package let citationMetadata: CitationMetadata?
|
|
|
|
/// Output only. Token count for this candidate.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Token count for this candidate.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// > Important: This property is not supported in the Gemini Enterprise Agent Platform.
|
|
package let tokenCount: Int?
|
|
|
|
/// Output only. Grounding metadata for the candidate.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Grounding metadata for the candidate.
|
|
///
|
|
/// This field is populated for `GenerateContent` calls.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. Metadata returned when grounding is enabled. It contains the
|
|
/// sources used to ground the generated content.
|
|
package let groundingMetadata: GroundingMetadata?
|
|
|
|
/// Output only. Average log probability score of the candidate.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Average log probability score of the candidate.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. The average log probability of the tokens in this candidate.
|
|
/// This is a length-normalized score that can be used to compare the quality
|
|
/// of candidates of different lengths. A higher average log probability
|
|
/// suggests a more confident and coherent response.
|
|
package let avgLogprobs: Double?
|
|
|
|
/// Output only. Log-likelihood scores for the response tokens and top tokens
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Log-likelihood scores for the response tokens and top tokens
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. The detailed log probability information for the tokens in
|
|
/// this candidate. This is useful for debugging, understanding model
|
|
/// uncertainty, and identifying potential "hallucinations".
|
|
package let logprobsResult: LogprobsResult?
|
|
|
|
/// Output only. Metadata related to url context retrieval tool.
|
|
///
|
|
/// ### Gemini Developer API
|
|
///
|
|
/// Output only. Metadata related to url context retrieval tool.
|
|
///
|
|
/// ### Gemini Enterprise Agent Platform
|
|
///
|
|
/// Output only. Metadata returned when the model uses the `url_context` tool
|
|
/// to get information from a user-provided URL.
|
|
package let urlContextMetadata: UrlContextMetadata?
|
|
|
|
/// Creates a new `Candidate`.
|
|
///
|
|
/// - Parameters:
|
|
/// - index: Output only. Index of the candidate in the list of response candidates. (behavior varies by backend). For more details, see ``index``.
|
|
/// - content: Output only. Generated content returned from the model. (behavior varies by backend). For more details, see ``content``.
|
|
/// - finishReason: Optional. Output only. The reason why the model stopped generating tokens. (behavior varies by backend). For more details, see ``finishReason``.
|
|
/// - finishMessage: Optional. Output only. Details the reason why the model stopped generating tokens. (behavior varies by backend). For more details, see ``finishMessage``.
|
|
/// - safetyRatings: List of ratings for the safety of a response candidate. (behavior varies by backend). For more details, see ``safetyRatings``.
|
|
/// - citationMetadata: Output only. Citation information for model-generated candidate. (behavior varies by backend). For more details, see ``citationMetadata``.
|
|
/// - tokenCount: Output only. Token count for this candidate. (Gemini Developer API only). For more details, see ``tokenCount``.
|
|
/// - groundingMetadata: Output only. Grounding metadata for the candidate. (behavior varies by backend). For more details, see ``groundingMetadata``.
|
|
/// - avgLogprobs: Output only. Average log probability score of the candidate. (behavior varies by backend). For more details, see ``avgLogprobs``.
|
|
/// - logprobsResult: Output only. Log-likelihood scores for the response tokens and top tokens (behavior varies by backend). For more details, see ``logprobsResult``.
|
|
/// - urlContextMetadata: Output only. Metadata related to url context retrieval tool. (behavior varies by backend). For more details, see ``urlContextMetadata``.
|
|
package init(
|
|
index: Int? = nil,
|
|
content: Content? = nil,
|
|
finishReason: FinishReason? = nil,
|
|
finishMessage: String? = nil,
|
|
safetyRatings: [SafetyRating]? = nil,
|
|
citationMetadata: CitationMetadata? = nil,
|
|
tokenCount: Int? = nil,
|
|
groundingMetadata: GroundingMetadata? = nil,
|
|
avgLogprobs: Double? = nil,
|
|
logprobsResult: LogprobsResult? = nil,
|
|
urlContextMetadata: UrlContextMetadata? = nil
|
|
) {
|
|
self.index = index
|
|
self.content = content
|
|
self.finishReason = finishReason
|
|
self.finishMessage = finishMessage
|
|
self.safetyRatings = safetyRatings
|
|
self.citationMetadata = citationMetadata
|
|
self.tokenCount = tokenCount
|
|
self.groundingMetadata = groundingMetadata
|
|
self.avgLogprobs = avgLogprobs
|
|
self.logprobsResult = logprobsResult
|
|
self.urlContextMetadata = urlContextMetadata
|
|
}
|
|
enum CodingKeys: String, CodingKey {
|
|
case index = "index"
|
|
case content = "content"
|
|
case finishReason = "finishReason"
|
|
case finishMessage = "finishMessage"
|
|
case safetyRatings = "safetyRatings"
|
|
case citationMetadata = "citationMetadata"
|
|
case tokenCount = "tokenCount"
|
|
case groundingMetadata = "groundingMetadata"
|
|
case avgLogprobs = "avgLogprobs"
|
|
case logprobsResult = "logprobsResult"
|
|
case urlContextMetadata = "urlContextMetadata"
|
|
}
|
|
}
|