OctoMed
/

OctoMed-7B

Image-Text-to-Text

text-generation-inference

Model card Files Files and versions

OctoMed commited on 7 days ago

Commit

02c01e3

·

verified ·

1 Parent(s): f8505ad

Update README

Files changed (1) hide show

README.md +2 -1

README.md CHANGED Viewed

@@ -284,8 +284,9 @@ Please reason step-by-step, and put your final answer within \\boxed{}.
 ### Known Issues
 * Model is sensitive to system prompt. We recommend using the default one.
 * The model is finetuned for multiple-choice VQA. The model may follow instructions for other tasks but is not extensively tested or post-trained to do so.
-* Multi-turn conversation tasks are not part of the SFT training, and therefore may not be logically coherent.
 ## Citation

 ### Known Issues
 * Model is sensitive to system prompt. We recommend using the default one.
 * The model is finetuned for multiple-choice VQA. The model may follow instructions for other tasks but is not extensively tested or post-trained to do so.
+* The model occasionally states that it cannot see the image even when the image is provided. Some of our text-only reasoning traces describe an image in words, and the expected reasoning trace indicates that the image is missing. Despite this, we observe high benchmark performance and the model can still reliably use visual information.
+We hope to address these concerns moving forward in future iterations!
 ## Citation