M-BERT Base ViT-B
Github Model Card
## Usage
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](https://github.com/FreddeFrallan/Multilingual-CLIP).
Once this is done, you can load and use the model with the following code
```python
from src import multilingual_clip
model = multilingual_clip.load_model('M-BERT-Base-ViT')
embeddings = model(['Älgen är skogens konung!', 'Wie leben Eisbären in der Antarktis?', 'Вы знали, что все белые медведи левши?'])
print(embeddings.shape)
# Yields: torch.Size([3, 640])
```
## About
A [BERT-base-multilingual](https://huggingface.co/bert-base-multilingual-cased) tuned to match the embedding space for [69 languages](https://github.com/FreddeFrallan/Multilingual-CLIP/blob/main/Model%20Cards/M-BERT%20Base%2069/Fine-Tune-Languages.md), to the embedding space of the CLIP text encoder which accompanies the ViT-B/32 vision encoder.
A full list of the 100 languages used during pre-training can be found [here](https://github.com/google-research/bert/blob/master/multilingual.md#list-of-languages), and a list of the 4069languages used during fine-tuning can be found in [SupportedLanguages.md](https://github.com/FreddeFrallan/Multilingual-CLIP/blob/main/Model%20Cards/M-BERT%20Base%2069/Fine-Tune-Languages.md).
Training data pairs was generated by sampling 40k sentences for each language from the combined descriptions of [GCC](https://ai.google.com/research/ConceptualCaptions/) + [MSCOCO](https://cocodataset.org/#home) + [VizWiz](https://vizwiz.org/tasks-and-datasets/image-captioning/), and translating them into the corresponding language.
All translation was done using the [AWS translate service](https://aws.amazon.com/translate/), the quality of these translations have currently not been analyzed, but one can assume the quality varies between the 69 languages.