|
|
|
|
|
|
by chrtng
608 days ago
|
|
|
Great question! Yes, GPT Driver runs according to the test prompt each time, which makes it resilient to small changes. To speed up execution, we also use a caching mechanism that runs quickly if nothing has changed, and only uses the models when needed. |
|
|