Collection/documents : few enhancements for clarity

[nishika26]

Describe the current behaviour
There are few of the components in these modules that holds the possibility of causing confusion among the users, list them here -

• when user hits the endpoint concerning listing collection(s), In the collection info api , the parameter include_url is defaulted to true since it is likely that clients would want the file urls to allow users to download. This makes the /document endpoint redundant because client can get all info in just one API call so they would always do that.
• the response body that collection creation gives back, gives us openai's vector store id in a parameter called "llm_service_id", this is too vague.
• when uploading document using upload document endpoint, above the target format parameter we see that it is written that it supports transformation of these file types - "pdf, docx, txt". our doc transformation only supports pdf to markdown transformation.

Describe the enhancement you'd like

• In case of the include_url parameter, it is better to default to false like other places.
• change the response parameter's name from "llm_service_id" and "llm_service_provider" to "knowledge_base_id" and "knowledge_base_provider"
• Update the docs which is accurate to what the platform is capable to taking and doing, more specifically, remove "docs, txt", from the line above the target format parameter in upload document endpoint.