Created new README to give users the way to create llm-d step by step

[ryojsb]

It's related to the following pull request.
llm-d/llm-d#124

I wanted step by step procedure after deploying llm-d with quickstart script, but there's no that document.

Therefore I referred to quickstart script and verified the steps.

Please let me know if you have question or if there's what I have to add such as mode descriptions and so on.

Thank you

[nerdalert]

@ryojsb Thanks for the contribution! Would you mind fixing up the MD lint here plx. We have an upcoming change to decompose the charts for better if you want to take a look here so the install will change some if interested in testing and helping over the next couple weeks. Can track the infra repo here Thanks!

CI: https://github.com/llm-d/llm-d-deployer/actions/runs/16187672595/job/45895967373?pr=367

markdownlint-cli2........................................................Failed
- hook id: markdownlint-cli2
- duration: 0.62s
- exit code: 1

markdownlint-cli2 v0.17.2 (markdownlint v0.37.4)
Finding: notes/testing-nixl-and-epp.md .github/PULL_REQUEST_TEMPLATE/pull_request_template.md quickstart/README-step-by-step.md CONTRIBUTING.md
Linting: 4 file(s)
Summary: 6 error(s)
quickstart/README-step-by-step.md:16 MD012/no-multiple-blanks Multiple consecutive blank lines [Expected: 1; Actual: 2]
quickstart/README-step-by-step.md:65:1 MD034/no-bare-urls Bare URL used [Context: "https://github.com/llm-d/llm-d..."]
quickstart/README-step-by-step.md:163 MD012/no-multiple-blanks Multiple consecutive blank lines [Expected: 1; Actual: 2]
quickstart/README-step-by-step.md:182 MD012/no-multiple-blanks Multiple consecutive blank lines [Expected: 1; Actual: 2]
quickstart/README-step-by-step.md:337 MD031/blanks-around-fences Fenced code blocks should be surrounded by blank lines [Context: "```"]
quickstart/README-step-by-step.md:338 MD031/blanks-around-fences Fenced code blocks should be surrounded by blank lines [Context: "```bash"]
markdownlint-cli2 v0.17.2 (markdownlint v0.37.4)
Finding: quickstart/examples/no-features/README.md quickstart/infra/openshift-aws.md README.md quickstart/README-minikube.md
Linting: 4 file(s)
Summary: 0 error(s)
markdownlint-cli2 v0.17.2 (markdownlint v0.37.4)
Finding: chart-dependencies/validating-deployment-and-testing.md quickstart/metrics-overview.md quickstart/README.md quickstart/grafana-setup.md
Linting: 4 file(s)
Summary: 0 error(s)

[ryojsb]

@nerdalert I really appreciated your reviwe and comment!
I modified the file.

I'm really interested in the test for the change of installation method from operator to helm!
Please let me know when I can start that test.
And I'm intended to change this Readme after switching the procedure.

Thank you

[nerdalert]

@ryojsb awesome, testing over the next couple of days would be much appreciated on the examples here https://github.com/llm-d-incubation/llm-d-infra/tree/main/quickstart/examples there is a small and large p/d examples. Also in the project Slack if you hit issues or want realtime chat if it helps with anything.

Typos and lint left in this for CI. Ty.

@@ -1,1 +1,1 @@
-Then, Apply CRDs for Gateway API Inference Extention.
+Then, Apply CRDs for Gateway API Inference Extension.
@@ -78,7 +78,7 @@ After that, deploy istiod.
-Then, Apply CRDs for Gateway API.
+Then, Apply CRDs for Gateway API.
@@ -119,7 +119,7 @@ helm upgrade -i \
-The resources are created as follows.␠
+The resources are created as follows.
@@ -325,7 +325,7 @@ spec:
-ModelService resources are created.␠
+ModelService resources are created.
@@ -339,4 +340,5 @@ llm-d-redis-master-5f77dd4bf9-4s5sp                        1/1     Running   0
 meta-llama-llama-3-2-3b-instruct-decode-6f5c75fc45-rbndl   2/2     Running   0          32s
 meta-llama-llama-3-2-3b-instruct-epp-6f5556dddd-x99s5      1/1     Running   0          32s
 meta-llama-llama-3-2-3b-instruct-prefill-d85997579-f7mts   1/1     Running   0          32s
-```
+\ No newline at end of file
+```

[ryojsb]

@nerdalert Thank you for your instruction.
I'll check it out and try it.

I've already joined into the slack group, so I'll let you know if I have some topic.

Thank you