Switch primary model to GPT-4 Turbo Preview with GPT-4 as fallback #435
Conversation
KennyDizi
changed the title
…turbo model is 3x and 2x cheaper respectively compared to the previous GPT-4 pricing
| git_provider="github" | ||
| publish_output=true | ||
| publish_output_progress=true | ||
| verbosity_level=0 # 0,1,2 | ||
| use_extra_bad_extensions=false | ||
| use_repo_settings_file=true | ||
| ai_timeout=180 | ||
| max_description_tokens = 500 | ||
| max_commits_tokens = 500 | ||
| max_description_tokens = 1000 |
There was a problem hiding this comment.
Increase max_description_tokens and max_commits_tokens because gpt-4 turbo model is 3x and 2x cheaper respectively compared to the previous GPT-4 pricing.
|
@KennyDizi We won't switch primary default model that quickly, needs more testing and feedback from the community It is a good idea to add support for this new model. You need to edit properly: |
|
and for the primary model, set: |
@mrT23 sure I've added support for gpt-4-1106-preview model |
|
@mrT23 let's keep this PR like this and wait for model update as well as feedback from community. |
|
Tested with model |
mrT23
added
invalid
|
gpt-3.5 is faster than both. Quality needs to be thoroughly tested to ensure a positive tradeoff when changing the default model. Support for the new model will be added here: |
PR Type:
Enhancement
PR Description:
This PR updates the primary model used by the application to "gpt-4-1106-preview" and sets "gpt-4" as the fallback model. The changes are made in the configuration file.
PR Main Files Walkthrough:
files:
pr_agent/settings/configuration.toml: The primary model has been changed from 'gpt-4' to 'gpt-4-1106-preview'. The fallback model has been updated from 'gpt-3.5-turbo-16k' to 'gpt-4'.