|
| 1 | +// Copyright 2024 Google LLC |
| 2 | +// |
| 3 | +// Licensed under the Apache License, Version 2.0 (the "License"); |
| 4 | +// you may not use this file except in compliance with the License. |
| 5 | +// You may obtain a copy of the License at |
| 6 | +// |
| 7 | +// http://www.apache.org/licenses/LICENSE-2.0 |
| 8 | +// |
| 9 | +// Unless required by applicable law or agreed to in writing, software |
| 10 | +// distributed under the License is distributed on an "AS IS" BASIS, |
| 11 | +// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 12 | +// See the License for the specific language governing permissions and |
| 13 | +// limitations under the License. |
| 14 | + |
| 15 | +import FirebaseCore |
| 16 | +import FirebaseVertexAI |
| 17 | +import XCTest |
| 18 | + |
| 19 | +#if canImport(UIKit) |
| 20 | + import UIKit |
| 21 | +#endif // canImport(UIKit) |
| 22 | + |
| 23 | +// These snippet tests are intentionally skipped in CI jobs; see the README file in this directory |
| 24 | +// for instructions on running them manually. |
| 25 | + |
| 26 | +@available(iOS 15.0, macOS 12.0, macCatalyst 15.0, tvOS 15.0, watchOS 8.0, *) |
| 27 | +final class MultimodalSnippets: XCTestCase { |
| 28 | + let bundle = BundleTestUtil.bundle() |
| 29 | + lazy var model = VertexAI.vertexAI().generativeModel(modelName: "gemini-1.5-flash") |
| 30 | + lazy var videoURL = { |
| 31 | + guard let url = bundle.url(forResource: "animals", withExtension: "mp4") else { |
| 32 | + fatalError("Video file animals.mp4 not found in Resources.") |
| 33 | + } |
| 34 | + return url |
| 35 | + }() |
| 36 | + |
| 37 | + override func setUpWithError() throws { |
| 38 | + try FirebaseApp.configureDefaultAppForSnippets() |
| 39 | + } |
| 40 | + |
| 41 | + override func tearDown() async throws { |
| 42 | + await FirebaseApp.deleteDefaultAppForSnippets() |
| 43 | + } |
| 44 | + |
| 45 | + #if canImport(UIKit) |
| 46 | + func testMultimodalOneImageNonStreaming() async throws { |
| 47 | + guard let image = UIImage(systemName: "bicycle") else { fatalError() } |
| 48 | + |
| 49 | + // Provide a text prompt to include with the image |
| 50 | + let prompt = "What's in this picture?" |
| 51 | + |
| 52 | + // To generate text output, call generateContent and pass in the prompt |
| 53 | + let response = try await model.generateContent(image, prompt) |
| 54 | + print(response.text ?? "No text in response.") |
| 55 | + } |
| 56 | + |
| 57 | + func testMultimodalOneImageStreaming() async throws { |
| 58 | + guard let image = UIImage(systemName: "bicycle") else { fatalError() } |
| 59 | + |
| 60 | + // Provide a text prompt to include with the image |
| 61 | + let prompt = "What's in this picture?" |
| 62 | + |
| 63 | + // To stream generated text output, call generateContentStream and pass in the prompt |
| 64 | + let contentStream = try model.generateContentStream(image, prompt) |
| 65 | + for try await chunk in contentStream { |
| 66 | + if let text = chunk.text { |
| 67 | + print(text) |
| 68 | + } |
| 69 | + } |
| 70 | + } |
| 71 | + |
| 72 | + func testMultimodalMultiImagesNonStreaming() async throws { |
| 73 | + guard let image1 = UIImage(systemName: "car") else { fatalError() } |
| 74 | + guard let image2 = UIImage(systemName: "car.2") else { fatalError() } |
| 75 | + |
| 76 | + // Provide a text prompt to include with the images |
| 77 | + let prompt = "What's different between these pictures?" |
| 78 | + |
| 79 | + // To generate text output, call generateContent and pass in the prompt |
| 80 | + let response = try await model.generateContent(image1, image2, prompt) |
| 81 | + print(response.text ?? "No text in response.") |
| 82 | + } |
| 83 | + |
| 84 | + func testMultimodalMultiImagesStreaming() async throws { |
| 85 | + guard let image1 = UIImage(systemName: "car") else { fatalError() } |
| 86 | + guard let image2 = UIImage(systemName: "car.2") else { fatalError() } |
| 87 | + |
| 88 | + // Provide a text prompt to include with the images |
| 89 | + let prompt = "What's different between these pictures?" |
| 90 | + |
| 91 | + // To stream generated text output, call generateContentStream and pass in the prompt |
| 92 | + let contentStream = try model.generateContentStream(image1, image2, prompt) |
| 93 | + for try await chunk in contentStream { |
| 94 | + if let text = chunk.text { |
| 95 | + print(text) |
| 96 | + } |
| 97 | + } |
| 98 | + } |
| 99 | + #endif // canImport(UIKit) |
| 100 | + |
| 101 | + func testMultimodalVideoNonStreaming() async throws { |
| 102 | + // Provide the video as `Data` with the appropriate MIME type |
| 103 | + let video = try InlineDataPart(data: Data(contentsOf: videoURL), mimeType: "video/mp4") |
| 104 | + |
| 105 | + // Provide a text prompt to include with the video |
| 106 | + let prompt = "What is in the video?" |
| 107 | + |
| 108 | + // To generate text output, call generateContent with the text and video |
| 109 | + let response = try await model.generateContent(video, prompt) |
| 110 | + print(response.text ?? "No text in response.") |
| 111 | + } |
| 112 | + |
| 113 | + func testMultimodalVideoStreaming() async throws { |
| 114 | + // Provide the video as `Data` with the appropriate MIME type |
| 115 | + let video = try InlineDataPart(data: Data(contentsOf: videoURL), mimeType: "video/mp4") |
| 116 | + |
| 117 | + // Provide a text prompt to include with the video |
| 118 | + let prompt = "What is in the video?" |
| 119 | + |
| 120 | + // To stream generated text output, call generateContentStream with the text and video |
| 121 | + let contentStream = try model.generateContentStream(video, prompt) |
| 122 | + for try await chunk in contentStream { |
| 123 | + if let text = chunk.text { |
| 124 | + print(text) |
| 125 | + } |
| 126 | + } |
| 127 | + } |
| 128 | +} |
0 commit comments