CharleyXu

PR #760

@langchain4j, Please take the time to code review, thanks！

I encountered the same problem. When parsing a large pdf document, an error occurred. ` DocumentSplitter splitter = DocumentSplitters.recursive(2000, 200); EmbeddingStoreIngestor ingestor = EmbeddingStoreIngestor.builder() .documentSplitter(splitter) .embeddingModel(embeddingModel) .embeddingStore(embeddingStore) .build(); ingestor.ingest(document) `...

> @mylzj at the moment there are no real agents implemented, but AI Service with Tools is kind of "light agent". ReAct and similar algorithms are yet to be implemented....

[opseval.txt](https://github.com/langchain4j/langchain4j/files/14584078/opseval.txt) thanks, already uploaded. AllMiniLmL6V2EmbeddingModel can work normally, but after changing to BgeSmallZhEmbeddingModel, an error will be reported during execution.

I found that the implementation of spring-ai-transformers refers to the ai.djl.huggingface.tokenizers package. I hope it will help the problem. ```xml org.springframework.ai spring-ai-transformers 0.8.0 ``` ```xml ai.djl.huggingface tokenizers 0.26.0 compile ```

ok,no rush.

The solution is issue #759

> Hi, thanks a lot! I it related to #712 ? Yes, the problem to be solved by PR #712 is the same. In comparison, the amount of code changes...

I didn't realize at the time that #712 had solved the problem.