嵌入模型，使用多模态报错

[lanyanhua]

ai:
# 阿里灵积
dash-scope:
embedding:
options:
model: multimodal-embedding-v1

org.springframework.ai.retry.NonTransientAiException: 400 - {"code":"InvalidParameter","message":"url error, please check url！","request_id":"0bd93f80-2bfd-93eb-b54e-29647f87cb12"} at org.springframework.ai.autoconfigure.retry.SpringAiRetryAutoConfiguration$2.handleError(SpringAiRetryAutoConfiguration.java:95) ~[spring-ai-spring-boot-autoconfigure-1.0.0-M3.jar:1.0.0-M3] at org.springframework.web.client.ResponseErrorHandler.handleError(ResponseErrorHandler.java:63) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.client.StatusHandler.lambda$fromErrorHandler$1(StatusHandler.java:71) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.client.StatusHandler.handle(StatusHandler.java:146) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.client.DefaultRestClient$DefaultResponseSpec.applyStatusHandlers(DefaultRestClient.java:698) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.client.DefaultRestClient.readWithMessageConverters(DefaultRestClient.java:200) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.client.DefaultRestClient$DefaultResponseSpec.readBody(DefaultRestClient.java:685) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.client.DefaultRestClient$DefaultResponseSpec.toEntityInternal(DefaultRestClient.java:655) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.client.DefaultRestClient$DefaultResponseSpec.toEntity(DefaultRestClient.java:644) ~[spring-web-6.1.11.jar:6.1.11] at com.alibaba.cloud.ai.dashscope.api.DashScopeApi.embeddings(DashScopeApi.java:294) ~[spring-ai-alibaba-core-1.0.0-M3.3.jar:1.0.0-M3.3] at com.alibaba.cloud.ai.dashscope.embedding.DashScopeEmbeddingModel.lambda$call$1(DashScopeEmbeddingModel.java:136) ~[spring-ai-alibaba-core-1.0.0-M3.3.jar:1.0.0-M3.3] at org.springframework.retry.support.RetryTemplate.doExecute(RetryTemplate.java:344) ~[spring-retry-2.0.7.jar:na] at org.springframework.retry.support.RetryTemplate.execute(RetryTemplate.java:217) ~[spring-retry-2.0.7.jar:na] at com.alibaba.cloud.ai.dashscope.embedding.DashScopeEmbeddingModel.lambda$call$3(DashScopeEmbeddingModel.java:133) ~[spring-ai-alibaba-core-1.0.0-M3.3.jar:1.0.0-M3.3] at io.micrometer.observation.Observation.observe(Observation.java:565) ~[micrometer-observation-1.13.2.jar:1.13.2] at com.alibaba.cloud.ai.dashscope.embedding.DashScopeEmbeddingModel.call(DashScopeEmbeddingModel.java:132) ~[spring-ai-alibaba-core-1.0.0-M3.3.jar:1.0.0-M3.3] at org.springframework.ai.embedding.EmbeddingModel.embed(EmbeddingModel.java:65) ~[spring-ai-core-1.0.0-M3.jar:1.0.0-M3] at org.springframework.ai.embedding.EmbeddingModel.embed(EmbeddingModel.java:47) ~[spring-ai-core-1.0.0-M3.jar:1.0.0-M3] at org.springframework.ai.embedding.AbstractEmbeddingModel.dimensions(AbstractEmbeddingModel.java:56) ~[spring-ai-core-1.0.0-M3.jar:1.0.0-M3] at org.springframework.ai.embedding.AbstractEmbeddingModel.dimensions(AbstractEmbeddingModel.java:78) ~[spring-ai-core-1.0.0-M3.jar:1.0.0-M3] at org.springframework.ai.vectorstore.MilvusVectorStore.createObservationContextBuilder(MilvusVectorStore.java:556) ~[spring-ai-milvus-store-1.0.0-M3.jar:1.0.0-M3] at org.springframework.ai.vectorstore.observation.AbstractObservationVectorStore.add(AbstractObservationVectorStore.java:51) ~[spring-ai-core-1.0.0-M3.jar:1.0.0-M3] at com.enshudata.esreport.ai.controller.VectorStoreController.addImages(VectorStoreController.java:133) ~[classes/:na] at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[na:na] at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77) ~[na:na] at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[na:na] at java.base/java.lang.reflect.Method.invoke(Method.java:568) ~[na:na] at org.springframework.web.method.support.InvocableHandlerMethod.doInvoke(InvocableHandlerMethod.java:255) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.method.support.InvocableHandlerMethod.invokeForRequest(InvocableHandlerMethod.java:188) ~[spring-web-6.1.11.jar:6.1.11] at org.springframework.web.servlet.mvc.method.annotation.ServletInvocableHandlerMethod.invokeAndHandle(ServletInvocableHandlerMethod.java:118) ~[spring-webmvc-6.1.11.jar:6.1.11] at org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerAdapter.invokeHandlerMethod(RequestMappingHandlerAdapter.java:926) ~[spring-webmvc-6.1.11.jar:6.1.11] at org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerAdapter.handleInternal(RequestMappingHandlerAdapter.java:831) ~[spring-webmvc-6.1.11.jar:6.1.11] at org.springframework.web.servlet.mvc.method.AbstractHandlerMethodAdapter.handle(AbstractHandlerMethodAdapter.java:87) ~[spring-webmvc-6.1.11.jar:6.1.11] at org.springframework.web.servlet.DispatcherServlet.doDispatch(DispatcherServlet.java:1089) ~[spring-webmvc-6.1.11.jar:6.1.11] at org.springframework.web.servlet.DispatcherServlet.doService(DispatcherServlet.java:979) ~[spring-webmvc-6.1.11.jar:6.1.11] at org.springframework.web.servlet.FrameworkServlet.processRequest(FrameworkServlet.java:1014) ~[spring-webmvc-6.1.11.jar:6.1.11] at org.springframework.web.servlet.FrameworkServlet.doPost(FrameworkServlet.java:914) ~[spring-webmvc-6.1.11.jar:6.1.11]

[PolarishT]

current spring ai alibaba does not support multimodal-embedding-v1 which is probably due to the fact that the multimodal-embedding-v1 URL is not the same as text-embedding.
But if you want to have a feature test, you can change DashscopeApi.embedding method.like this

MutimodelEmbeddingRequest request = new MutimodelEmbeddingRequest(List.of(objectObjectHashMap).toString(), "multimodal-embedding-v1");
		return this.restClient.post()
			.uri("/api/v1/services/embeddings/multimodal-embedding/multimodal-embedding")
			.body(request)
			.retrieve()
			.toEntity(EmbeddingList.class);

and test method like this

var dashScopeApi = new DashScopeApi(System.getenv("DASHSCOPE_API_KEY"));
		var embeddingModel = new DashScopeEmbeddingModel(dashScopeApi, MetadataMode.EMBED,
				DashScopeEmbeddingOptions.builder()
						.withModel("multimodal-embedding-v1")
						.build());

		EmbeddingResponse embeddingResponse = embeddingModel.embedForResponse(List.of("hello world"));

		assertThat(embeddingResponse.getResults()).hasSize(1);
		assertThat(embeddingResponse.getResults().get(0)).isNotNull();
		assertThat(embeddingResponse.getResults().get(0).getOutput()).hasSize(1536);
		Assertions.assertEquals(embeddingResponse.getMetadata().get("model"), "multimodal-embedding-v1");

[lanyanhua]

current spring ai alibaba does not support multimodal-embedding-v1 which is probably due to the fact that the multimodal-embedding-v1 URL is not the same as text-embedding. But if you want to have a feature test, you can change DashscopeApi.embedding method.like this

MutimodelEmbeddingRequest request = new MutimodelEmbeddingRequest(List.of(objectObjectHashMap).toString(), "multimodal-embedding-v1");
return this.restClient.post()
.uri("/api/v1/services/embeddings/multimodal-embedding/multimodal-embedding")
.body(request)
.retrieve()
.toEntity(EmbeddingList.class);
and test method like this

var dashScopeApi = new DashScopeApi(System.getenv("DASHSCOPE_API_KEY"));
var embeddingModel = new DashScopeEmbeddingModel(dashScopeApi, MetadataMode.EMBED,
DashScopeEmbeddingOptions.builder()
.withModel("multimodal-embedding-v1")
.build());

  EmbeddingResponse embeddingResponse = embeddingModel.embedForResponse(List.of("hello world"));

  assertThat(embeddingResponse.getResults()).hasSize(1);
  assertThat(embeddingResponse.getResults().get(0)).isNotNull();
  assertThat(embeddingResponse.getResults().get(0).getOutput()).hasSize(1536);
  Assertions.assertEquals(embeddingResponse.getMetadata().get("model"), "multimodal-embedding-v1");

代码可以贴全一点吗，第一个 MutimodelEmbeddingRequest 找不到类。第二个报错 {"code":"InvalidParameter","message":"url error, please check url！","request_id":"6add799d-8890-9a4f-b391-2947203720cf"}

[lanyanhua]

current spring ai alibaba does not support multimodal-embedding-v1 which is probably due to the fact that the multimodal-embedding-v1 URL is not the same as text-embedding. But if you want to have a feature test, you can change DashscopeApi.embedding method.like this
MutimodelEmbeddingRequest request = new MutimodelEmbeddingRequest(List.of(objectObjectHashMap).toString(), "multimodal-embedding-v1");
return this.restClient.post()
.uri("/api/v1/services/embeddings/multimodal-embedding/multimodal-embedding")
.body(request)
.retrieve()
.toEntity(EmbeddingList.class);
and test method like this
var dashScopeApi = new DashScopeApi(System.getenv("DASHSCOPE_API_KEY"));
var embeddingModel = new DashScopeEmbeddingModel(dashScopeApi, MetadataMode.EMBED,
DashScopeEmbeddingOptions.builder()
.withModel("multimodal-embedding-v1")
.build());

  EmbeddingResponse embeddingResponse = embeddingModel.embedForResponse(List.of("hello world"));

  assertThat(embeddingResponse.getResults()).hasSize(1);
  assertThat(embeddingResponse.getResults().get(0)).isNotNull();
  assertThat(embeddingResponse.getResults().get(0).getOutput()).hasSize(1536);
  Assertions.assertEquals(embeddingResponse.getMetadata().get("model"), "multimodal-embedding-v1");

代码可以贴全一点吗，第一个 MutimodelEmbeddingRequest 找不到类。第二个报错 {"code":"InvalidParameter","message":"url error, please check url！","request_id":"6add799d-8890-9a4f-b391-2947203720cf"}

参考接口写https://help.aliyun.com/zh/model-studio/developer-reference/multimodal-embedding-api-reference?spm=a2c4g.11186623.help-menu-2400256.d_3_3_19.41244d4d12sNnE
第二个方式还是没好