Update zero-shot table to match results reported in paper.

README.md

@@ -121,7 +121,7 @@ We also use a weight decay of 0.2 and a batch size of 32K.
 ### Testing Data
 
 We tested BioCLIP on the following collection of 10 biologically-relevant tasks.
- - [Meta-Album](https://paperswithcode.com/dataset/meta-album): Specifically, we used the Plankton, Insects, Insects 2, PlantNet, Fungi, PlantVillage, Medicinal Leaf, and PlantDoc datasets from Set-0 through Set-2 (Set-3 was still not released as of our publication/evaluation (Nov. 2023).
+ - [Meta-Album](https://paperswithcode.com/dataset/meta-album): Specifically, we used the Plankton, Insects, Insects 2, PlantNet, Fungi, PlantVillage, Medicinal Leaf, and PlantDoc datasets from Set-0 through Set-2 (Set-3 was still not released as of our publication/evaluation (Nov. 2023)).
  - [Birds 525](https://www.kaggle.com/datasets/gpiosenka/100-bird-species): We evaluated on the 2,625 test images provided with the dataset.
  - [Rare Species](https://huggingface.co/datasets/imageomics/rare-species): A new dataset we curated for the purpose of testing this model and to contribute to the ML for Conservation community. It consists of 400 species labeled Near Threatened through Extinct in the Wild by the [IUCN Red List](https://www.iucnredlist.org/), with 30 images per species. For more information, see our dataset, [Rare Species](https://huggingface.co/datasets/imageomics/rare-species).
 
@@ -170,8 +170,8 @@ Please see [our paper](https://doi.org/10.48550/arXiv.2311.18803) for few-shot r
       <td>5.4</td>
       <td>15.9</td>
       <td>26.1</td>
-      <td>26.6</td>
-      <td>21.4</td>
+      <td>31.8</td>
+      <td>21.9</td>
     </tr>
     <tr>
       <td>OpenCLIP</td>
@@ -184,36 +184,36 @@ Please see [our paper](https://doi.org/10.48550/arXiv.2311.18803) for few-shot r
       <td>8.0</td>
       <td>12.4</td>
       <td>25.8</td>
-      <td>31.0</td>
-      <td>20.6</td>
+      <td>29.8</td>
+      <td>20.4</td>
     </tr>
     <tr>
       <td>BioCLIP</td>
-      <td><b>74.7</b></td>
-      <td><b>5.4</b></td>
-      <td><b>32.7</b></td>
-      <td><b>21.2</b></td>
-      <td><b>91.0</b></td>
-      <td><b>51.8</b></td>
-      <td><b>24.0</b></td>
-      <td><b>48.1</b></td>
-      <td><b>27.5</b></td>
-      <td><b>39.2</b></td>
-      <td><b>41.5</b></td>
+      <td><b>72.1</b></td>
+      <td><b>6.1</b></td>
+      <td><b>34.8</b></td>
+      <td><b>20.4</b></td>
+      <td><b>91.4</b></td>
+      <td>40.7</td>
+      <td><b>24.4</b></td>
+      <td><b>38.6</b></td>
+      <td><b>28.4</b></td>
+      <td><b>38.0</b></td>
+      <td><b>39.4</b></td>
     </tr>
     <tr>
       <td>iNat21 Only</td>
-      <td>55.7</td>
-      <td>2.7</td>
-      <td>29.9</td>
-      <td>12.0</td>
-      <td>89.3</td>
-      <td>42.7</td>
-      <td>16.4</td>
-      <td>22.2</td>
-      <td>18.8</td>
-      <td>19.4</td>
-      <td>30.9</td>
+      <td>56.1</td>
+      <td>2.6</td>
+      <td>30.7</td>
+      <td>11.5</td>
+      <td>88.2</td>
+      <td><b>43.0</b></td>
+      <td>18.4</td>
+      <td>25.6</td>
+      <td>20.5</td>
+      <td>21.3</td>
+      <td>31.7</td>
     </tr>
   </tbody>
 </table>
@@ -221,7 +221,7 @@ Please see [our paper](https://doi.org/10.48550/arXiv.2311.18803) for few-shot r
 
 ### Summary
 
-BioCLIP outperforms general-domain baselines by 18% on average.
+BioCLIP outperforms general-domain baselines by 17% on average for zero-shot.
 
 ### Model Examination