M-CLIP/Swedish-2M

To use this model along with the original CLIP vision encoder you need to download the code and additional linear weights from the Multilingual-CLIP Github Github model . The model is tuned to match the embedding space of the CLIP text encoder which accompanies the Res50x4 vision encoding . Training data pairs were generated by sampling 2 Million sentences from the combined descriptions of GCC + MSCOCO + VizWiz, and translating them into Swedish . The Huggingface Opus Model was done using the HuggingFace Model, which seemingly procudes higher quality translations than relying on the AWS translate service . To use the model with the model, you can load and use the,