Is your feature request related to a problem? Please describe.
I have created two profiles for different models/api keys. One is for OpenAI and the other is for Azure.
When adding the --os flag, it executes this code:
if args.os:
args.profile = "os.py"
This does not allow me to switch between credentials without re-writing os.py every time I want to switch.
Describe the solution you'd like
When I run OS Mode, I would like to be able to choose which credentials are used.
The command for this would be interpreter -p profile1.yaml --os and would use values from both files. Any conflicting fields would have profile1.yaml overwrite the values in os.py. I believe the only conflict would usually be llm.model
Describe alternatives you've considered
No response
Additional context
The .py profiles are more intimidating to edit than the .yaml profiles. Most users would only want to edit the yaml profiles so they should be able to control most configurations from there.
Is your feature request related to a problem? Please describe.
I have created two profiles for different models/api keys. One is for OpenAI and the other is for Azure.
When adding the
--os
flag, it executes this code:This does not allow me to switch between credentials without re-writing
os.py
every time I want to switch.Describe the solution you'd like
When I run
OS Mode
, I would like to be able to choose which credentials are used.The command for this would be
interpreter -p profile1.yaml --os
and would use values from both files. Any conflicting fields would haveprofile1.yaml
overwrite the values inos.py
. I believe the only conflict would usually bellm.model
Describe alternatives you've considered
No response
Additional context
The
.py
profiles are more intimidating to edit than the.yaml
profiles. Most users would only want to edit the yaml profiles so they should be able to control most configurations from there.