This is the feature alignment pre-training work to train only only the multi-modal projector. "Predict" paragraph given caption, ocr and image token
- Downloads last month
- 2
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.