Update llm inference docs

[alabulei1]

fix the build errors

[alabulei1]

Hello, I am a code review bot on flows.network. Here are my reviews of code commits in this PR.

---

Commit 7b602846faeae24d6100f68f07f3fb89fce2b2c9

Key changes in the patch:

1. In the English version of the documentation:

   • The link to the Llama2 inference in Rust chapter has been updated to use the correct file name ("llm_inference" instead of "llm-inference").
   • Minor formatting changes have been made in the installation commands.

2. In the Chinese version of the documentation:

   • The link to the Llama2 inference in Rust chapter has been updated to use the correct file name ("llm_inference" instead of "llm-inference").
   • Additional information has been added for the ggml-backend and its installation steps.

Potential problems:

1. The links to the Llama2 inference in Rust chapter have been updated, but other chapters' links have not been checked for accuracy. It would be good to verify that all hyperlinks are correct.
2. The formatting changes in the installation commands may introduce errors if not properly reviewed.
3. In the Chinese version, there may be a typo in the heading "具有 TensorFlow-Lite 后端的 WASI-NN 插件" (third-level heading "带有 TensorFlow-Lite 后端的 WASI-NN 插件") where "后端" is repeated twice.
4. The patch only includes changes to the documentation files, so it's unclear if any corresponding code changes were made to support these updates.