Open mtoniott opened 8 months ago
This feature requested very often (e.g. #14). Initially, I thought it is better to separate user from underlying models so they can only focus on the efficiency provided by tlm and went with codellama:7b as the most optimal solution. But I'm planning to include an option to go with models with less parameters like starcoder2:3b
.
Thank you for your feedback.
+1 on this, it is better to give users the ability to choose the desired model, but tlm
can set the system prompts ... otherwise with any new model release, users have to wait for the new tlm
release to use it (if you decide to change the default). for example, right now the more optimal model might be deepseek-coder-v2
or codestral
but there is no way to change it
This feature is a must have !
Not all computers can run codellama. Being able to select another model would be quite cool for those.