Official repository of OFA (ICML 2022). Paper: OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework
In pretraining/finetuning stage, for vision-language task (especially for visual_grounding and caption), can I set the length of generated tokens? Because I want a longer generated sentence.
Also for captioning evaluation, how can I set the output length?
I found there are some arguments like max-length-a and max-length-b. Could you help me understand the meaning?
Dear coauthors,
In pretraining/finetuning stage, for vision-language task (especially for visual_grounding and caption), can I set the length of generated tokens? Because I want a longer generated sentence.
Also for captioning evaluation, how can I set the output length?
I found there are some arguments like
max-length-a
andmax-length-b
. Could you help me understand the meaning?Thanks a lot!