Often when I using this plugin, it's because I want to verify how fast certain models are.
The benchmarking doesn't seem to be tracked separately for separate models as of yet.
It's ONE AVG TPS and TFFT for the conversation.
When I switch models, it would be nice if:
- Opencode switches back the metrics to what was previously measured for this specific model.
- OR I get the a command or something to reset the tracking so I can benchmark this model fresh.
Often when I using this plugin, it's because I want to verify how fast certain models are.
The benchmarking doesn't seem to be tracked separately for separate models as of yet.
It's ONE AVG TPS and TFFT for the conversation.
When I switch models, it would be nice if: