chore(llm): using nuitka to provide a binary/perf way for the service

[weijinglin]

follow #199 to build a new docker image using Nuitka in Dockerfile.nk

• Only simple tests have been conducted so far
• more complete tests are still needed

close #200

[Copilot]

Pull Request Overview

This PR introduces a new Docker image build process using Nuitka for a binary/embed release along with enhancements to batch text embedding methods across multiple embedding providers.

• Removed threaded processing in favor of a batched embedding API call in build_semantic_index.py
• Added new get_texts_embeddings methods with documentation in embedding provider implementations (qianfan, openai, ollama)
• Updated dependency versions for Ollama in requirements.txt and pyproject.toml, and added a new Dockerfile.nk for the build process

Reviewed Changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
hugegraph_llm/operators/index_op/build_semantic_index.py	Removed ThreadPoolExecutor and switched to using batched embedding API
hugegraph_llm/models/embeddings/qianfan.py	Added batched embedding method with new docstring
hugegraph_llm/models/embeddings/openai.py	Added batched embedding method with comprehensive docstring
hugegraph_llm/models/embeddings/ollama.py	Updated API usage for embeddings and added batched embedding method
hugegraph_llm/models/embeddings/base.py	Added abstract method for batched embedding to maintain consistency
hugegraph-llm/requirements.txt, pyproject.toml	Updated Ollama dependency version
docker/Dockerfile.nk	Added new Dockerfile using Nuitka for binary/embed release

Comments suppressed due to low confidence (3)

hugegraph_llm/models/embeddings/qianfan.py:55

• [nitpick] Consider revising the docstring phrasing for clarity, e.g., changing 'Usage refer:' to 'See:' or 'Usage:' to align with standard documentation practices.

        """ Usage refer: https://cloud.baidu.com/doc/WENXINWORKSHOP/s/hlmokk9qn"""

hugegraph_llm/operators/index_op/build_semantic_index.py:40

• The change from using ThreadPoolExecutor to a direct batched API call assumes that get_texts_embeddings scales well with larger lists. Please verify that this approach performs adequately for high-volume inputs.

embeddings = self.embedding.get_texts_embeddings(vids)

hugegraph_llm/models/embeddings/ollama.py:43

• The update from self.client.embeddings to self.client.embed reflects the new API version. Please ensure that the parameters and expected return structure are fully aligned with the updated Ollama documentation.

return list(self.client.embed(model=self.model, input=text)["embeddings"][0])

[weijinglin]

The product of image constructed by Dockerfile.nk

[imbajin]

better to attach a tree structure for it

merge it & add docs for it later