when asking for certain pictures, the generated request was not the best. this new version changes things to make it more understandable for the LLM as well as including important context.
After testing in my own environment, and is working correctly for all types or prompts.
Currently working on getting different image types (for example, if you same "can you send a close up photo of 'x'" it will inject "close up" into the prompt to the LLM.
when asking for certain pictures, the generated request was not the best. this new version changes things to make it more understandable for the LLM as well as including important context.
After testing in my own environment, and is working correctly for all types or prompts.
Currently working on getting different image types (for example, if you same "can you send a close up photo of 'x'" it will inject "close up" into the prompt to the LLM.