fix: CLI conveniences (add-on to #674)

[cpacker]

Please describe the purpose of this pull request

Improving the workflow a bit more.

• Only raise errors about openai or azure keys missing once the user selects openai/azure as an option
  • Do this for both the LLM endpoint step and the embeddings endpoint step
    • e.g. if the user selects localllm and no openai key is specified, don't throw an error until they try to use openai for embeddings
  • For openai, allow the user to provide the key in real-time (if missing)
• When key-related errors happen in configure, they're thrown as ValueErrors
  • Catch the ValueErrors and print them cleanly vs a fat stacktrace

---

Tested: example runs of memgpt configure

User does not have openai key set, and tries to do openai LLM

% memgpt configure
? Select LLM inference provider: openai
? Enter your OpenAI API key (starts with 'sk-', see https://platform.openai.com/api-
keys): banana
? Override default endpoint: https://api.openai.com/v1
? Select default model (recommended: gpt-4): gpt-4
? Select embedding provider: openai
? Select default preset: memgpt_chat
? Select default persona: sam_pov
? Select default human: basic
? Select storage backend for archival data: local
Saving config to /Users/loaner/.memgpt/config

User does not have openai key set, tried to do local LLM + openai embeddings

 % memgpt configure
? Select LLM inference provider: local
? Select LLM backend (select 'openai' if you have an OpenAI compatible proxy): lmstu
dio
? Enter default endpoint: http://localhost:1234
? Select default model wrapper (recommended: chatml): chatml
? Select your model's context window (for Mistral 7B models, this is probably 8k / 8
192): 8192
? Select embedding provider: openai
? Enter your OpenAI API key (starts with 'sk-', see https://platform.openai.com/api-
keys): banana
? Select default preset: memgpt_chat
? Select default persona: sam_pov
? Select default human: basic
? Select storage backend for archival data: local
Saving config to /Users/loaner/.memgpt/config

User has no config, but openai is set in the env vars

% memgpt configure            
? Select LLM inference provider: openai
? Override default endpoint: https://api.openai.com/v1
? Select default model (recommended: gpt-4): gpt-4
? Select embedding provider: openai
? Select default preset: memgpt_chat
? Select default persona: sam_pov
? Select default human: basic
? Select storage backend for archival data: local
Saving config to /Users/loaner/.memgpt/config

User has no azure vars set, tries to use azure LLM (immediate error)

 % memgpt configure            
? Select LLM inference provider: azure
Missing environment variables for Azure (see https://memgpt.readme.io/docs/endpoints#azure-openai). Please set then run `memgpt configure` again.

User has no azure vars set, tries to use local LLM but azure embedding

 % memgpt configure
? Select LLM inference provider: local
? Select LLM backend (select 'openai' if you have an OpenAI compatible proxy): lmstu
dio
? Enter default endpoint: http://localhost:1234
? Select default model wrapper (recommended: chatml): chatml
? Select your model's context window (for Mistral 7B models, this is probably 8k / 8
192): 8192
? Select embedding provider: azure
Missing environment variables for Azure (see https://memgpt.readme.io/docs/endpoints#azure-openai). Please set then run `memgpt configure` again.

User has only azure LLM vars set, tries to use azure LLM + embeddings (expected success, since in get_azure_credentials we default from LLM->embeddings)

% memgpt configure
? Select LLM inference provider: azure
? Select default model (recommended: gpt-4): gpt-4-1106-preview
? Select embedding provider: azure
? Select default preset: memgpt_chat
? Select default persona: sam_pov
? Select default human: basic
? Select storage backend for archival data: local
📖 Saving config to /Users/loaner/.memgpt/config

User has only azure embedding vars set, tries to use both LLM + embeddings

% memgpt configure
? Select LLM inference provider: azure
Missing environment variables for Azure (see https://memgpt.readme.io/docs/endpoints#azure-openai). Please set then run `memgpt configure` again.

[cpacker]

@sarahwooders is this the correct way to save the key back to the file in "real-time" if the user provides it here?