Add fairness evals for Idefics instruct
Browse files
README.md
CHANGED
@@ -301,6 +301,20 @@ Idefics Instruct Evaluations:
|
|
301 |
| | 16 | 66.8 | 51.7 | 31.6 | 44.8 | 70.2 | 128.8 | 101.5 | 75.8 | - | 51.7 | - | 63.3 | - |
|
302 |
| | 32 | 66.9 | 52.3 | 32.0 | 46.0 | 71.7 | 127.8 | 101.0 | 76.3 | - | 50.8 | - | 60.9 | - |
|
303 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
304 |
IDEFICS vs IDEFICS-instruct.
|
305 |
| Model | Shots | <nobr>VQAv2<br>OE VQA acc.</nobr> | <nobr>OKVQA<br>OE VQA acc.</nobr> | <nobr>TextVQA<br>OE VQA acc.</nobr> | <nobr>VizWiz<br>OE VQA acc.</nobr> | <nobr>TextCaps<br>CIDEr</nobr> | <nobr>Coco<br>CIDEr</nobr> | <nobr>NoCaps<br>CIDEr</nobr> | <nobr>Flickr<br>CIDEr</nobr> | <nobr>VisDial<br>NDCG</nobr> | <nobr>HatefulMemes<br>ROC AUC</nobr> | <nobr>ScienceQA<br>acc.</nobr> | <nobr>RenderedSST2<br>acc.</nobr> | <nobr>Winoground<br>group (text/image)</nobr> |
|
306 |
|:----------------------------------------|:--------|---------------------:|---------------------:|-----------------------:|----------------------:|-------------------:|---------------:|-----------------:|-----------------:|-----------------:|-------------------------:|-----------------------:|--------------------------:|----------------------------------:|
|
|
|
301 |
| | 16 | 66.8 | 51.7 | 31.6 | 44.8 | 70.2 | 128.8 | 101.5 | 75.8 | - | 51.7 | - | 63.3 | - |
|
302 |
| | 32 | 66.9 | 52.3 | 32.0 | 46.0 | 71.7 | 127.8 | 101.0 | 76.3 | - | 50.8 | - | 60.9 | - |
|
303 |
|
304 |
+
Fairness Evaluations:
|
305 |
+
| Model | Shots | <nobr>FairFaceGender<br>acc.</nobr> | <nobr>FairFaceRace<br>acc.</nobr> | <nobr>FairFaceAge<br>acc.</nobr> |
|
306 |
+
|:---------------------|--------:|----------------------------:|--------------------------:|-------------------------:|
|
307 |
+
| 80B IDEFICS Instruct | 0 | 95.7 | 63.4 | 47.1 |
|
308 |
+
| | 4 | 95.6 | 51.4 | 48.3 |
|
309 |
+
| | 8 | 95.8 | 51.0 | 51.1 |
|
310 |
+
| | 16 | 96.1 | 47.6 | 51.8 |
|
311 |
+
| | 32 | 96.2 | 36.8 | 50.3 |
|
312 |
+
| 9B IDEFICS Instruct | 0 | 92.7 | 59.6 | 43.9 |
|
313 |
+
| | 4 | 95.2 | 43.3 | 38.7 |
|
314 |
+
| | 8 | 95.8 | 51.7 | 40.1 |
|
315 |
+
| | 16 | 96.1 | 58.9 | 41.7 |
|
316 |
+
| | 32 | 96.1 | 59.7 | 44.8 |
|
317 |
+
|
318 |
IDEFICS vs IDEFICS-instruct.
|
319 |
| Model | Shots | <nobr>VQAv2<br>OE VQA acc.</nobr> | <nobr>OKVQA<br>OE VQA acc.</nobr> | <nobr>TextVQA<br>OE VQA acc.</nobr> | <nobr>VizWiz<br>OE VQA acc.</nobr> | <nobr>TextCaps<br>CIDEr</nobr> | <nobr>Coco<br>CIDEr</nobr> | <nobr>NoCaps<br>CIDEr</nobr> | <nobr>Flickr<br>CIDEr</nobr> | <nobr>VisDial<br>NDCG</nobr> | <nobr>HatefulMemes<br>ROC AUC</nobr> | <nobr>ScienceQA<br>acc.</nobr> | <nobr>RenderedSST2<br>acc.</nobr> | <nobr>Winoground<br>group (text/image)</nobr> |
|
320 |
|:----------------------------------------|:--------|---------------------:|---------------------:|-----------------------:|----------------------:|-------------------:|---------------:|-----------------:|-----------------:|-----------------:|-------------------------:|-----------------------:|--------------------------:|----------------------------------:|
|