ollama - Please provide bettern README or fix the bug

[magzim21]

The readme says "Set PROVIDER in your environment to ollama"

$ export PROVIDER=ollama
$ export AI_PROVIDER=ollama # tried this too
$ ai-commit -h                                  
Please set the OPENAI_API_KEY environment variable.

[mbenapari]

Hi @magzim21 ,

I've submitted a pull request to address this issue. Hopefully, it will be merged soon!
Feel free to take a look: PR #35.

[sidsarasvati]

thanks for the issue, yes same, pretty useless README on how to use ollama
The PR seems to be open for over a month now

@insulineru ?

[sidsarasvati]

Anyway - this helped resolved my issue -
#32

[miguel-san-martin]

ai-commit --PROVIDER=ollama --MODEL=mistral

[pu-007]

ai-commit --PROVIDER=ollama --MODEL=mistral

need install from git first
#34 (comment)

[mlab817]

I also found that adding MODEL to the environment variables should set the model automatically allowing you to run:

MODEL=mistral

ai-commit

without specifying variables. The command ai-commit is looking for gpt 4.0 mini model

[gagareg]

I set all the variables and arguments, but it didn’t help.

[/t/git] (00:02:44) ❯❯❯ echo $PROVIDER $AI_PROVIDER $MODEL
ollama ollama mistral
[/t/git] (00:03:46) ❯❯❯ ai-commit --PROVIDER=ollama --MODEL=mistral
Please set the OPENAI_API_KEY environment variable.