Update README.md

README.md

@@ -12,7 +12,9 @@ tags:
 
 # CapDec - NoiseLevel: 0.015
 
-This are model weights originally provided by the authors of the paper [Text-Only Training for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf).
+## Model Description
+
+These are model weights originally provided by the authors of the paper [Text-Only Training for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf).
 
 Their method aims to train CLIP with only text samples. Therefore they are injecting zero-mean Gaussian Noise into the text embeddings before decoding.
 
@@ -28,7 +30,9 @@ The "Noise Level" of 0.015 is equivalent to the Noise Variance which is the squa
 The reported metrics are results of a model with a Noise Variance of 0.016, which the authors unfortunately do not provide in their repository. 
 This model with a Noise Variance 0.015 is the closest available  pre-trained model to their best model.
 
+## Datasets
+The authors trained the model on MS-COCO and Flickr30k datasets.
 
 ## Performance
 The authors don't explicitly report the performance for this NoiseLevel but it can be estimated from the following figure from the original paper:
-
+![](capdec_performance.png)