tuman
/

vit-rugpt2-image-captioning

vision-encoder-decoder

image-captioning

Inference Endpoints

Model card Files Files and versions Community

tuman commited on Jan 25, 2023

Commit

13342bb

•

1 Parent(s): 5e52925

Update README.md

Files changed (1) hide show

README.md +5 -1

README.md CHANGED Viewed

@@ -9,10 +9,14 @@ metrics:
 library_name: transformers
 ---
-# vit-rugpt2-image-captioning
 This is an image captioning model trained on translated version (en-ru) of dataset COCO2014.
 # Metrics on test data
 * Bleu: 8.672

 library_name: transformers
 ---
+# First image captioning model for russian language vit-rugpt2-image-captioning
 This is an image captioning model trained on translated version (en-ru) of dataset COCO2014.
+# Model Details
+Model was initialized `sberbank-ai/rugpt3large_based_on_gpt2` for encoder and `google/vit-base-patch16-224-in21k` for decoder.
 # Metrics on test data
 * Bleu: 8.672