Hugging Face's logo

Spaces:
gzomer
/
clip-multilingual
Runtime error

App Files Community
gzomer commited on
Commit
34a1e3c
•
1 Parent(s): 2cc278d

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +1 -1
app.py CHANGED
@@ -25,7 +25,7 @@ semantic_search = load_model()
25
  description = '''
26
  # Multilingual Semantic Search
27
 
28
- **Search images in 100 languages (list [here](https://github.com/pytorch/fairseq/blob/main/examples/xlmr/README.md#introduction)) powered by MultiLingual CLIP**
29
 
30
  MultiLingual CLIP is a custom model built using OpenAI's [CLIP](https://openai.com/blog/clip/) and [XMLRoBERTa](https://huggingface.co/xlm-roberta-base) models, trained using 16 [Habana](https://habana.ai/) accelerators with PyTorch Lightning, Distributed Data Parallel, Mixed precision and using [COCO](https://cocodataset.org/) and [Google Conceptual Captions](https://ai.google.com/research/ConceptualCaptions) as training datasets.
31
 
25
  description = '''
26
  # Multilingual Semantic Search
27
 
28
+ **Search images in 100 languages (list [here](https://github.com/pytorch/fairseq/blob/main/examples/xlmr/README.md#introduction)) powered by [MultiLingual CLIP](https://huggingface.co/gzomer/clip-multilingual).**
29
 
30
  MultiLingual CLIP is a custom model built using OpenAI's [CLIP](https://openai.com/blog/clip/) and [XMLRoBERTa](https://huggingface.co/xlm-roberta-base) models, trained using 16 [Habana](https://habana.ai/) accelerators with PyTorch Lightning, Distributed Data Parallel, Mixed precision and using [COCO](https://cocodataset.org/) and [Google Conceptual Captions](https://ai.google.com/research/ConceptualCaptions) as training datasets.
31