but i found gemma-e4b is still too "dumb", and barely capable to provide any good response.
could you share your experience with how you use e2b to generate good result?
for screenshot generation its not open ended generation, image ocr and windows title is fed..and only structured json is asked in response, it works fine
so i just designed around it - small model + tight prompt + real context instead of hoping the model is clever. what were you trying to generate? can share the exact prompt setup if it helps