g8a9 commited on
Commit
d16ed94
1 Parent(s): 8f903f5

Add static rendering of tables.

Browse files
introduction.md CHANGED
@@ -193,6 +193,8 @@ described by the original caption. As evaluation metrics we use the MRR@K.
193
  | MRR@5 | **0.5039** | 0.3957|
194
  | MRR@10 | **0.5204** | 0.4129|
195
 
 
 
196
  It is true that we used the training set of MSCOCO-IT in training, and this might give us an advantage. However, the original CLIP model was trained
197
  on 400million images (and some of them might have been from MSCOCO).
198
 
@@ -209,6 +211,8 @@ We evaluate the models computing the accuracy at different levels.
209
  | Accuracy@10 | **52.55** | 42.91 |
210
  | Accuracy@100 | **81.08** | 67.11 |
211
 
 
 
212
  ### Discussion
213
 
214
  Our results confirm that CLIP-Italian is very competitive and beats mCLIP on the two different task
 
193
  | MRR@5 | **0.5039** | 0.3957|
194
  | MRR@10 | **0.5204** | 0.4129|
195
 
196
+ _If the table above doesn not show, you can have a look at it [here](https://huggingface.co/spaces/clip-italian/clip-italian-demo/raw/main/static/img/table_imagenet.png)._
197
+
198
  It is true that we used the training set of MSCOCO-IT in training, and this might give us an advantage. However, the original CLIP model was trained
199
  on 400million images (and some of them might have been from MSCOCO).
200
 
 
211
  | Accuracy@10 | **52.55** | 42.91 |
212
  | Accuracy@100 | **81.08** | 67.11 |
213
 
214
+ _If the table above doesn not show, you can have a look at it [here](https://huggingface.co/spaces/clip-italian/clip-italian-demo/raw/main/static/img/table_IR.png)._
215
+
216
  ### Discussion
217
 
218
  Our results confirm that CLIP-Italian is very competitive and beats mCLIP on the two different task
static/img/table_IR.png ADDED
static/img/table_imagenet.png ADDED