🐛🔧 #197: Handle max_tokens & content_filtered finish reasons across OpenAI, Azure and Cohere

[roma-glushko]

There is only one stream finish_reason handled as of now which is complete. But others are possible too.

Changes

In this PR, I'm handling the finish reason related to:

• text length bound passed in request params
• content moderation/toxicity
• Cohere streaming workflow doesn't seem to be working as errMapper was not really initialized. I have fixed that in this PR
• Cohere now ignores stream chunk types that Glide doesn't support like citation related stuff
• Cohere stream chunks are not set with the correct model name (e.g. some placeholder was used before)

[codecov]

Codecov Report

Attention: Patch coverage is 53.39806% with 48 lines in your changes are missing coverage. Please review.

Project coverage is 65.98%. Comparing base (12df04f) to head (2d0cad3).

Files	Patch %	Lines
pkg/providers/cohere/finish_reason.go	16.00%	21 Missing ⚠️
pkg/providers/cohere/chat_stream.go	44.00%	14 Missing ⚠️
pkg/providers/openai/finish_reasons.go	56.00%	11 Missing ⚠️
pkg/providers/cohere/chat.go	0.00%	1 Missing ⚠️
pkg/routers/config.go	0.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #194      +/-   ##
===========================================
+ Coverage    65.69%   65.98%   +0.29%     
===========================================
  Files           74       76       +2     
  Lines         3518     3546      +28     
===========================================
+ Hits          2311     2340      +29     
+ Misses        1082     1081       -1     
  Partials       125      125              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.