feat: Getting embeddings using text-embedding-005. (GoogleCloudPlatform#9603)

lee1premium-2024 · web-flow · commit d295d370464b · 2024-10-28T12:19:49.000-07:00
diff --git a/aiplatform/src/main/java/aiplatform/EmbeddingBatchSample.java b/aiplatform/src/main/java/aiplatform/EmbeddingBatchSample.java
@@ -40,7 +40,7 @@ public static void main(String[] args) throws IOException, InterruptedException
     // Could be a BigQuery table or a Google Cloud Storage file.
     // E.g. "gs://[BUCKET]/[OUTPUT].jsonl" OR "bq://[PROJECT].[DATASET].[TABLE]"
     String outputUri = "gs://YOUR_BUCKET/embedding_batch_output";
-    String textEmbeddingModel = "textembedding-gecko@003";
+    String textEmbeddingModel = "text-embedding-005";
 
     embeddingBatchSample(project, location, inputUri, outputUri, textEmbeddingModel);
   }
diff --git a/aiplatform/src/main/java/aiplatform/EmbeddingModelTuningSample.java b/aiplatform/src/main/java/aiplatform/EmbeddingModelTuningSample.java
@@ -84,7 +84,7 @@ public static PipelineJob createEmbeddingModelTuningPipelineJob(
     Matcher matcher = Pattern.compile("^(?<Location>\\w+-\\w+)").matcher(apiEndpoint);
     String location = matcher.matches() ? matcher.group("Location") : "us-central1";
     String templateUri =
-        "https://us-kfp.pkg.dev/ml-pipeline/llm-text-embedding/tune-text-embedding-model/v1.1.3";
+        "https://us-kfp.pkg.dev/ml-pipeline/llm-text-embedding/tune-text-embedding-model/v1.1.4";
     PipelineServiceSettings settings =
         PipelineServiceSettings.newBuilder().setEndpoint(apiEndpoint).build();
     try (PipelineServiceClient client = PipelineServiceClient.create(settings)) {
diff --git a/aiplatform/src/main/java/aiplatform/PredictTextEmbeddingsSample.java b/aiplatform/src/main/java/aiplatform/PredictTextEmbeddingsSample.java
@@ -41,7 +41,7 @@ public static void main(String[] args) throws IOException {
     // https://cloud.google.com/vertex-ai/docs/generative-ai/embeddings/get-text-embeddings
     String endpoint = "us-central1-aiplatform.googleapis.com:443";
     String project = "YOUR_PROJECT_ID";
-    String model = "text-embedding-004";
+    String model = "text-embedding-005";
     predictTextEmbeddings(
         endpoint,
         project,
diff --git a/aiplatform/src/main/java/aiplatform/PredictTextEmbeddingsSamplePreview.java b/aiplatform/src/main/java/aiplatform/PredictTextEmbeddingsSamplePreview.java
@@ -41,7 +41,7 @@ public static void main(String[] args) throws IOException {
     // https://cloud.google.com/vertex-ai/docs/generative-ai/embeddings/get-text-embeddings
     String endpoint = "us-central1-aiplatform.googleapis.com";
     String project = "YOUR_PROJECT_ID";
-    String model = "text-embedding-preview-0815";
+    String model = "text-embedding-005";
     // Calculate the embedding for a code retrieval query. Using 'CODE_RETRIEVAL_QUERY' for query.
     predictTextEmbeddings(
         endpoint,
diff --git a/aiplatform/src/test/java/aiplatform/EmbeddingBatchSampleTest.java b/aiplatform/src/test/java/aiplatform/EmbeddingBatchSampleTest.java
@@ -41,7 +41,7 @@ public class EmbeddingBatchSampleTest extends TestCase {
       "gs://cloud-samples-data/generative-ai/embeddings/embeddings_input.jsonl";
   private static final String GCS_OUTPUT_URI =
       String.format("gs://%s/embedding_batch_output", BUCKET_NAME);
-  private static final String MODEL_ID = "textembedding-gecko@003";
+  private static final String MODEL_ID = "text-embedding-005";
   static Storage storage;
   static Bucket bucket;
 
diff --git a/aiplatform/src/test/java/aiplatform/EmbeddingModelTuningSampleTest.java b/aiplatform/src/test/java/aiplatform/EmbeddingModelTuningSampleTest.java
@@ -53,7 +53,7 @@ public class EmbeddingModelTuningSampleTest {
 
   private static final String API_ENDPOINT = "us-central1-aiplatform.googleapis.com:443";
   private static final String PROJECT = System.getenv("UCAIP_PROJECT_ID");
-  private static final String BASE_MODEL_VERSION_ID = "text-embedding-004";
+  private static final String BASE_MODEL_VERSION_ID = "text-embedding-005";
   private static final String TASK_TYPE = "DEFAULT";
   private static final String JOB_DISPLAY_NAME = "embedding-customization-pipeline-sample";
   private static final String CORPUS = 
diff --git a/aiplatform/src/test/java/aiplatform/PredictTextEmbeddingsSampleTest.java b/aiplatform/src/test/java/aiplatform/PredictTextEmbeddingsSampleTest.java
@@ -52,7 +52,7 @@ public void testPredictTextEmbeddings() throws IOException {
         PredictTextEmbeddingsSample.predictTextEmbeddings(
             APIS_ENDPOINT,
             PROJECT,
-            "text-embedding-004",
+            "text-embedding-005",
             texts,
             "QUESTION_ANSWERING",
             OptionalInt.of(5));
@@ -70,7 +70,7 @@ public void testPredictTextEmbeddingsPreview() throws IOException {
         PredictTextEmbeddingsSamplePreview.predictTextEmbeddings(
             APIS_ENDPOINT,
             PROJECT,
-            "text-embedding-preview-0815",
+            "text-embedding-005",
             texts,
             "CODE_RETRIEVAL_QUERY",
             OptionalInt.of(5));

Original file line number	Diff line number	Diff line change
`@@ -40,7 +40,7 @@ public static void main(String[] args) throws IOException, InterruptedException`
`40`	`40`	`// Could be a BigQuery table or a Google Cloud Storage file.`
`41`	`41`	`// E.g. "gs://[BUCKET]/[OUTPUT].jsonl" OR "bq://[PROJECT].[DATASET].[TABLE]"`
`42`	`42`	`String outputUri = "gs://YOUR_BUCKET/embedding_batch_output";`
`43`		`- String textEmbeddingModel = "textembedding-gecko@003";`
	`43`	`+ String textEmbeddingModel = "text-embedding-005";`
`44`	`44`
`45`	`45`	`embeddingBatchSample(project, location, inputUri, outputUri, textEmbeddingModel);`
`46`	`46`	`}`