[Example] Add llama streaming example

[dm4]

• Should be merged after wasi-nn plugin update
• Depends on PR https://github.com/WasmEdge/WasmEdge/pull/2944

[juntao]

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

---

Overall Summary: The pull request adds a new example for llama streaming using WASI-NN and GGML as the backend. However, there are some potential issues and errors that need to be addressed. The instructions provided in the README.md file may be missing some steps or additional dependencies/setup instructions. The code in the main.rs file also seems to be incomplete, as it is cut off at the end. Additionally, there is a change in the --nn-preload flag where the architecture value is changed from CPU to AUTO without clear justification. The commit message should provide more context on the reason for this change and its impact.

Important Findings:

• Instructions and completeness of the README.md file need to be verified.
• Completeness of the code in the main.rs file needs to be checked.
• Justification for the change in the --nn-preload flag from CPU to AUTO should be provided in the commit message or comments.

Details

Commit 858fc0ce7835206d0c9dbaaa7e1639d87534a38c

Key changes:

• Added a new directory called "wasmedge-ggml-llama-interactive-stream".
• Added a new file called "Cargo.toml" with dependencies and package information.
• Added a new file called "README.md" with instructions and information about the project.
• Added a new file called "main.rs" with the main code for the llama streaming example.

Potential problems:

• There might be some missing or incomplete steps in the instructions provided in the README.md file, such as additional dependencies or setup instructions.
• The code in the main.rs file might have some missing or incomplete parts, as it is cut off at the end.

Overall, the changes seem to add a new example for llama streaming using WASI-NN and GGML as the backend. However, further review is needed to ensure the completeness and correctness of the additions.

Commit 0a89c8c3b7e32e4e2ed2fc31b2fde2d6f0665d75

Key changes:

• The --nn-preload flag in the wasmedge command has been modified.
• The architecture value in the --nn-preload flag has been changed from CPU to AUTO.

Potential problems:

• It is not clear why the architecture value was changed from CPU to AUTO in the --nn-preload flag. This change should be justified in the pull request description or comments.
• The commit message should provide more context on why this change was made and how it impacts the functionality or performance.

[hydai]

This is replaced by #52